{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.988968202465931,
  "eval_steps": 100.0,
  "global_step": 1925,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025957170668397143,
      "grad_norm": 64.5,
      "learning_rate": 5.181347150259068e-07,
      "loss": 0.6761971116065979,
      "memory(GiB)": 82.76,
      "step": 1,
      "token_acc": 0.8360655737704918,
      "train_speed(iter/s)": 0.019472
    },
    {
      "epoch": 0.012978585334198572,
      "grad_norm": 66.0,
      "learning_rate": 2.5906735751295338e-06,
      "loss": 0.6650555729866028,
      "memory(GiB)": 118.47,
      "step": 5,
      "token_acc": 0.8231644260599793,
      "train_speed(iter/s)": 0.021208
    },
    {
      "epoch": 0.025957170668397145,
      "grad_norm": 48.5,
      "learning_rate": 5.1813471502590676e-06,
      "loss": 0.4840095043182373,
      "memory(GiB)": 118.47,
      "step": 10,
      "token_acc": 0.8467274233637117,
      "train_speed(iter/s)": 0.020876
    },
    {
      "epoch": 0.03893575600259572,
      "grad_norm": 137.0,
      "learning_rate": 7.772020725388602e-06,
      "loss": 0.3320765018463135,
      "memory(GiB)": 118.47,
      "step": 15,
      "token_acc": 0.8696356275303644,
      "train_speed(iter/s)": 0.021309
    },
    {
      "epoch": 0.05191434133679429,
      "grad_norm": 21.25,
      "learning_rate": 1.0362694300518135e-05,
      "loss": 0.3111098766326904,
      "memory(GiB)": 118.47,
      "step": 20,
      "token_acc": 0.8665008291873963,
      "train_speed(iter/s)": 0.02131
    },
    {
      "epoch": 0.06489292667099286,
      "grad_norm": 21.5,
      "learning_rate": 1.2953367875647668e-05,
      "loss": 0.24997963905334472,
      "memory(GiB)": 118.47,
      "step": 25,
      "token_acc": 0.8876127973748975,
      "train_speed(iter/s)": 0.021488
    },
    {
      "epoch": 0.07787151200519143,
      "grad_norm": 25.5,
      "learning_rate": 1.5544041450777204e-05,
      "loss": 0.21027085781097413,
      "memory(GiB)": 118.47,
      "step": 30,
      "token_acc": 0.9171974522292994,
      "train_speed(iter/s)": 0.021184
    },
    {
      "epoch": 0.09085009733939,
      "grad_norm": 19.25,
      "learning_rate": 1.813471502590674e-05,
      "loss": 0.18175365924835205,
      "memory(GiB)": 118.47,
      "step": 35,
      "token_acc": 0.9322742474916388,
      "train_speed(iter/s)": 0.021197
    },
    {
      "epoch": 0.10382868267358858,
      "grad_norm": 42.0,
      "learning_rate": 2.072538860103627e-05,
      "loss": 0.2070298671722412,
      "memory(GiB)": 118.47,
      "step": 40,
      "token_acc": 0.9324546952224053,
      "train_speed(iter/s)": 0.021199
    },
    {
      "epoch": 0.11680726800778715,
      "grad_norm": 31.0,
      "learning_rate": 2.3316062176165805e-05,
      "loss": 0.14612390995025634,
      "memory(GiB)": 118.47,
      "step": 45,
      "token_acc": 0.9431356620633631,
      "train_speed(iter/s)": 0.021162
    },
    {
      "epoch": 0.12978585334198572,
      "grad_norm": 33.5,
      "learning_rate": 2.5906735751295337e-05,
      "loss": 0.1252074122428894,
      "memory(GiB)": 118.47,
      "step": 50,
      "token_acc": 0.9554140127388535,
      "train_speed(iter/s)": 0.02116
    },
    {
      "epoch": 0.1427644386761843,
      "grad_norm": 14.6875,
      "learning_rate": 2.8497409326424872e-05,
      "loss": 0.08470200896263122,
      "memory(GiB)": 118.47,
      "step": 55,
      "token_acc": 0.9732360097323601,
      "train_speed(iter/s)": 0.021224
    },
    {
      "epoch": 0.15574302401038287,
      "grad_norm": 24.75,
      "learning_rate": 3.108808290155441e-05,
      "loss": 0.1468730688095093,
      "memory(GiB)": 118.47,
      "step": 60,
      "token_acc": 0.9489539748953975,
      "train_speed(iter/s)": 0.021111
    },
    {
      "epoch": 0.16872160934458144,
      "grad_norm": 13.375,
      "learning_rate": 3.367875647668394e-05,
      "loss": 0.0989870309829712,
      "memory(GiB)": 118.47,
      "step": 65,
      "token_acc": 0.9626016260162602,
      "train_speed(iter/s)": 0.021069
    },
    {
      "epoch": 0.18170019467878,
      "grad_norm": 29.25,
      "learning_rate": 3.626943005181348e-05,
      "loss": 0.15612951517105103,
      "memory(GiB)": 118.47,
      "step": 70,
      "token_acc": 0.9571428571428572,
      "train_speed(iter/s)": 0.021002
    },
    {
      "epoch": 0.1946787800129786,
      "grad_norm": 16.0,
      "learning_rate": 3.886010362694301e-05,
      "loss": 0.10940381288528442,
      "memory(GiB)": 118.47,
      "step": 75,
      "token_acc": 0.9690635451505016,
      "train_speed(iter/s)": 0.02104
    },
    {
      "epoch": 0.20765736534717716,
      "grad_norm": 11.5,
      "learning_rate": 4.145077720207254e-05,
      "loss": 0.13801568746566772,
      "memory(GiB)": 118.47,
      "step": 80,
      "token_acc": 0.9573378839590444,
      "train_speed(iter/s)": 0.020886
    },
    {
      "epoch": 0.22063595068137573,
      "grad_norm": 28.375,
      "learning_rate": 4.404145077720208e-05,
      "loss": 0.11146789789199829,
      "memory(GiB)": 118.47,
      "step": 85,
      "token_acc": 0.9662447257383966,
      "train_speed(iter/s)": 0.020723
    },
    {
      "epoch": 0.2336145360155743,
      "grad_norm": 8.125,
      "learning_rate": 4.663212435233161e-05,
      "loss": 0.1379382848739624,
      "memory(GiB)": 118.47,
      "step": 90,
      "token_acc": 0.9566284779050737,
      "train_speed(iter/s)": 0.020619
    },
    {
      "epoch": 0.24659312134977288,
      "grad_norm": 11.9375,
      "learning_rate": 4.922279792746114e-05,
      "loss": 0.13294199705123902,
      "memory(GiB)": 118.47,
      "step": 95,
      "token_acc": 0.9616985845129059,
      "train_speed(iter/s)": 0.020597
    },
    {
      "epoch": 0.25957170668397145,
      "grad_norm": 78.5,
      "learning_rate": 5.1813471502590674e-05,
      "loss": 0.15107507705688478,
      "memory(GiB)": 118.47,
      "step": 100,
      "token_acc": 0.9649415692821369,
      "train_speed(iter/s)": 0.020542
    },
    {
      "epoch": 0.27255029201817,
      "grad_norm": 13.75,
      "learning_rate": 5.440414507772021e-05,
      "loss": 0.27871944904327395,
      "memory(GiB)": 118.47,
      "step": 105,
      "token_acc": 0.9241034195162635,
      "train_speed(iter/s)": 0.02047
    },
    {
      "epoch": 0.2855288773523686,
      "grad_norm": 7.1875,
      "learning_rate": 5.6994818652849744e-05,
      "loss": 0.14314144849777222,
      "memory(GiB)": 118.47,
      "step": 110,
      "token_acc": 0.9567027477102414,
      "train_speed(iter/s)": 0.020365
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 7.40625,
      "learning_rate": 5.9585492227979276e-05,
      "loss": 0.14128265380859376,
      "memory(GiB)": 118.47,
      "step": 115,
      "token_acc": 0.9611248966087675,
      "train_speed(iter/s)": 0.02028
    },
    {
      "epoch": 0.31148604802076574,
      "grad_norm": 7.625,
      "learning_rate": 6.217616580310881e-05,
      "loss": 0.15114243030548097,
      "memory(GiB)": 118.47,
      "step": 120,
      "token_acc": 0.9528145695364238,
      "train_speed(iter/s)": 0.02023
    },
    {
      "epoch": 0.3244646333549643,
      "grad_norm": 15.875,
      "learning_rate": 6.476683937823834e-05,
      "loss": 0.14014118909835815,
      "memory(GiB)": 118.47,
      "step": 125,
      "token_acc": 0.9568755085435313,
      "train_speed(iter/s)": 0.020211
    },
    {
      "epoch": 0.3374432186891629,
      "grad_norm": 13.0625,
      "learning_rate": 6.735751295336788e-05,
      "loss": 0.14676700830459594,
      "memory(GiB)": 118.47,
      "step": 130,
      "token_acc": 0.9623430962343096,
      "train_speed(iter/s)": 0.020168
    },
    {
      "epoch": 0.3504218040233614,
      "grad_norm": 11.5625,
      "learning_rate": 6.994818652849742e-05,
      "loss": 0.15908313989639283,
      "memory(GiB)": 118.47,
      "step": 135,
      "token_acc": 0.9551495016611296,
      "train_speed(iter/s)": 0.020164
    },
    {
      "epoch": 0.36340038935756,
      "grad_norm": 8.375,
      "learning_rate": 7.253886010362695e-05,
      "loss": 0.14115771055221557,
      "memory(GiB)": 118.47,
      "step": 140,
      "token_acc": 0.953757225433526,
      "train_speed(iter/s)": 0.02013
    },
    {
      "epoch": 0.37637897469175857,
      "grad_norm": 12.5,
      "learning_rate": 7.512953367875648e-05,
      "loss": 0.17947659492492676,
      "memory(GiB)": 118.47,
      "step": 145,
      "token_acc": 0.947986577181208,
      "train_speed(iter/s)": 0.020045
    },
    {
      "epoch": 0.3893575600259572,
      "grad_norm": 12.6875,
      "learning_rate": 7.772020725388602e-05,
      "loss": 0.13564352989196776,
      "memory(GiB)": 118.47,
      "step": 150,
      "token_acc": 0.9545454545454546,
      "train_speed(iter/s)": 0.020024
    },
    {
      "epoch": 0.4023361453601557,
      "grad_norm": 9.375,
      "learning_rate": 8.031088082901554e-05,
      "loss": 0.14768315553665162,
      "memory(GiB)": 118.47,
      "step": 155,
      "token_acc": 0.954248366013072,
      "train_speed(iter/s)": 0.02003
    },
    {
      "epoch": 0.4153147306943543,
      "grad_norm": 8.5625,
      "learning_rate": 8.290155440414508e-05,
      "loss": 0.15876761674880982,
      "memory(GiB)": 118.47,
      "step": 160,
      "token_acc": 0.955701394585726,
      "train_speed(iter/s)": 0.020035
    },
    {
      "epoch": 0.42829331602855286,
      "grad_norm": 5.6875,
      "learning_rate": 8.549222797927462e-05,
      "loss": 0.20421419143676758,
      "memory(GiB)": 118.47,
      "step": 165,
      "token_acc": 0.9363484087102177,
      "train_speed(iter/s)": 0.020018
    },
    {
      "epoch": 0.44127190136275146,
      "grad_norm": 8.4375,
      "learning_rate": 8.808290155440416e-05,
      "loss": 0.13153988122940063,
      "memory(GiB)": 118.47,
      "step": 170,
      "token_acc": 0.9637860082304527,
      "train_speed(iter/s)": 0.02002
    },
    {
      "epoch": 0.45425048669695,
      "grad_norm": 8.375,
      "learning_rate": 9.067357512953368e-05,
      "loss": 0.3276111364364624,
      "memory(GiB)": 118.47,
      "step": 175,
      "token_acc": 0.9358245329000813,
      "train_speed(iter/s)": 0.020018
    },
    {
      "epoch": 0.4672290720311486,
      "grad_norm": 12.1875,
      "learning_rate": 9.326424870466322e-05,
      "loss": 0.13705794811248778,
      "memory(GiB)": 118.47,
      "step": 180,
      "token_acc": 0.9646090534979423,
      "train_speed(iter/s)": 0.020066
    },
    {
      "epoch": 0.48020765736534715,
      "grad_norm": 10.0,
      "learning_rate": 9.585492227979275e-05,
      "loss": 0.17206543684005737,
      "memory(GiB)": 118.47,
      "step": 185,
      "token_acc": 0.9485903814262023,
      "train_speed(iter/s)": 0.020072
    },
    {
      "epoch": 0.49318624269954575,
      "grad_norm": 65.5,
      "learning_rate": 9.844559585492228e-05,
      "loss": 0.15847107172012329,
      "memory(GiB)": 118.47,
      "step": 190,
      "token_acc": 0.9510135135135135,
      "train_speed(iter/s)": 0.020079
    },
    {
      "epoch": 0.5061648280337443,
      "grad_norm": 6.75,
      "learning_rate": 9.999967099424576e-05,
      "loss": 0.12832045555114746,
      "memory(GiB)": 118.47,
      "step": 195,
      "token_acc": 0.9608482871125612,
      "train_speed(iter/s)": 0.020091
    },
    {
      "epoch": 0.5191434133679429,
      "grad_norm": 21.625,
      "learning_rate": 9.999596972923527e-05,
      "loss": 0.2224121570587158,
      "memory(GiB)": 118.47,
      "step": 200,
      "token_acc": 0.9467312348668281,
      "train_speed(iter/s)": 0.020136
    },
    {
      "epoch": 0.5321219987021415,
      "grad_norm": 105.5,
      "learning_rate": 9.99881562474689e-05,
      "loss": 0.3189336538314819,
      "memory(GiB)": 118.47,
      "step": 205,
      "token_acc": 0.9142156862745098,
      "train_speed(iter/s)": 0.020143
    },
    {
      "epoch": 0.54510058403634,
      "grad_norm": 14.0,
      "learning_rate": 9.997623119161306e-05,
      "loss": 0.3505317211151123,
      "memory(GiB)": 118.47,
      "step": 210,
      "token_acc": 0.9171452009844134,
      "train_speed(iter/s)": 0.020174
    },
    {
      "epoch": 0.5580791693705386,
      "grad_norm": 7.8125,
      "learning_rate": 9.996019554251512e-05,
      "loss": 0.17262498140335084,
      "memory(GiB)": 118.47,
      "step": 215,
      "token_acc": 0.9438110749185668,
      "train_speed(iter/s)": 0.020194
    },
    {
      "epoch": 0.5710577547047372,
      "grad_norm": 9.6875,
      "learning_rate": 9.994005061912266e-05,
      "loss": 0.19659049510955812,
      "memory(GiB)": 118.47,
      "step": 220,
      "token_acc": 0.9433805162364696,
      "train_speed(iter/s)": 0.020193
    },
    {
      "epoch": 0.5840363400389358,
      "grad_norm": 5.78125,
      "learning_rate": 9.99157980783751e-05,
      "loss": 0.1692946434020996,
      "memory(GiB)": 118.47,
      "step": 225,
      "token_acc": 0.9464138499587799,
      "train_speed(iter/s)": 0.020175
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 8.0625,
      "learning_rate": 9.98874399150673e-05,
      "loss": 0.15234674215316774,
      "memory(GiB)": 118.47,
      "step": 230,
      "token_acc": 0.9539094650205762,
      "train_speed(iter/s)": 0.020195
    },
    {
      "epoch": 0.6099935107073329,
      "grad_norm": 3.796875,
      "learning_rate": 9.985497846168557e-05,
      "loss": 0.16389089822769165,
      "memory(GiB)": 118.47,
      "step": 235,
      "token_acc": 0.9533333333333334,
      "train_speed(iter/s)": 0.020211
    },
    {
      "epoch": 0.6229720960415315,
      "grad_norm": 5.65625,
      "learning_rate": 9.981841638821576e-05,
      "loss": 0.1235802173614502,
      "memory(GiB)": 118.47,
      "step": 240,
      "token_acc": 0.9637972646822205,
      "train_speed(iter/s)": 0.020238
    },
    {
      "epoch": 0.6359506813757301,
      "grad_norm": 9.5,
      "learning_rate": 9.977775670192371e-05,
      "loss": 0.2110598564147949,
      "memory(GiB)": 118.47,
      "step": 245,
      "token_acc": 0.9404081632653061,
      "train_speed(iter/s)": 0.020262
    },
    {
      "epoch": 0.6489292667099286,
      "grad_norm": 4.34375,
      "learning_rate": 9.973300274710787e-05,
      "loss": 0.1242341160774231,
      "memory(GiB)": 118.47,
      "step": 250,
      "token_acc": 0.9559800664451827,
      "train_speed(iter/s)": 0.020264
    },
    {
      "epoch": 0.6619078520441272,
      "grad_norm": 32.25,
      "learning_rate": 9.968415820482415e-05,
      "loss": 1.1298725128173828,
      "memory(GiB)": 118.47,
      "step": 255,
      "token_acc": 0.8556193601312552,
      "train_speed(iter/s)": 0.020271
    },
    {
      "epoch": 0.6748864373783258,
      "grad_norm": 20.5,
      "learning_rate": 9.963122709258335e-05,
      "loss": 0.6246148586273194,
      "memory(GiB)": 118.47,
      "step": 260,
      "token_acc": 0.8447154471544716,
      "train_speed(iter/s)": 0.020304
    },
    {
      "epoch": 0.6878650227125244,
      "grad_norm": 596.0,
      "learning_rate": 9.957421376402051e-05,
      "loss": 0.5492882251739502,
      "memory(GiB)": 118.47,
      "step": 265,
      "token_acc": 0.9152404237978811,
      "train_speed(iter/s)": 0.020343
    },
    {
      "epoch": 0.7008436080467229,
      "grad_norm": 5.8125,
      "learning_rate": 9.951312290853691e-05,
      "loss": 0.1511799216270447,
      "memory(GiB)": 118.47,
      "step": 270,
      "token_acc": 0.9506688963210702,
      "train_speed(iter/s)": 0.02037
    },
    {
      "epoch": 0.7138221933809215,
      "grad_norm": 154.0,
      "learning_rate": 9.944795955091438e-05,
      "loss": 1.6441532135009767,
      "memory(GiB)": 118.47,
      "step": 275,
      "token_acc": 0.8274153592072667,
      "train_speed(iter/s)": 0.020369
    },
    {
      "epoch": 0.72680077871512,
      "grad_norm": 7.75,
      "learning_rate": 9.937872905090196e-05,
      "loss": 1.1411833763122559,
      "memory(GiB)": 118.47,
      "step": 280,
      "token_acc": 0.8543371522094927,
      "train_speed(iter/s)": 0.020393
    },
    {
      "epoch": 0.7397793640493187,
      "grad_norm": 13.6875,
      "learning_rate": 9.930543710277509e-05,
      "loss": 0.2547419548034668,
      "memory(GiB)": 118.47,
      "step": 285,
      "token_acc": 0.9494702526487367,
      "train_speed(iter/s)": 0.020415
    },
    {
      "epoch": 0.7527579493835171,
      "grad_norm": 27.75,
      "learning_rate": 9.922808973486721e-05,
      "loss": 0.21210157871246338,
      "memory(GiB)": 118.47,
      "step": 290,
      "token_acc": 0.9428571428571428,
      "train_speed(iter/s)": 0.020426
    },
    {
      "epoch": 0.7657365347177157,
      "grad_norm": 220.0,
      "learning_rate": 9.914669330907399e-05,
      "loss": 0.1741060733795166,
      "memory(GiB)": 118.47,
      "step": 295,
      "token_acc": 0.9474116680361545,
      "train_speed(iter/s)": 0.020432
    },
    {
      "epoch": 0.7787151200519143,
      "grad_norm": 11.5,
      "learning_rate": 9.906125452032999e-05,
      "loss": 0.19344786405563355,
      "memory(GiB)": 118.47,
      "step": 300,
      "token_acc": 0.928453947368421,
      "train_speed(iter/s)": 0.020448
    },
    {
      "epoch": 0.791693705386113,
      "grad_norm": 3.109375,
      "learning_rate": 9.897178039605802e-05,
      "loss": 0.20231380462646484,
      "memory(GiB)": 118.47,
      "step": 305,
      "token_acc": 0.9286898839137645,
      "train_speed(iter/s)": 0.020452
    },
    {
      "epoch": 0.8046722907203114,
      "grad_norm": 4.59375,
      "learning_rate": 9.887827829559117e-05,
      "loss": 0.16842379570007324,
      "memory(GiB)": 118.47,
      "step": 310,
      "token_acc": 0.9444444444444444,
      "train_speed(iter/s)": 0.020467
    },
    {
      "epoch": 0.81765087605451,
      "grad_norm": 2.875,
      "learning_rate": 9.878075590956742e-05,
      "loss": 0.1668983817100525,
      "memory(GiB)": 118.47,
      "step": 315,
      "token_acc": 0.9469026548672567,
      "train_speed(iter/s)": 0.020482
    },
    {
      "epoch": 0.8306294613887086,
      "grad_norm": 4.40625,
      "learning_rate": 9.867922125929712e-05,
      "loss": 0.15300171375274657,
      "memory(GiB)": 118.47,
      "step": 320,
      "token_acc": 0.9525395503746877,
      "train_speed(iter/s)": 0.020508
    },
    {
      "epoch": 0.8436080467229072,
      "grad_norm": 3.3125,
      "learning_rate": 9.857368269610324e-05,
      "loss": 0.16267441511154174,
      "memory(GiB)": 118.47,
      "step": 325,
      "token_acc": 0.9539094650205762,
      "train_speed(iter/s)": 0.020505
    },
    {
      "epoch": 0.8565866320571057,
      "grad_norm": 7.21875,
      "learning_rate": 9.846414890063446e-05,
      "loss": 0.1828203558921814,
      "memory(GiB)": 118.47,
      "step": 330,
      "token_acc": 0.9417637271214643,
      "train_speed(iter/s)": 0.020508
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 4.53125,
      "learning_rate": 9.835062888215114e-05,
      "loss": 0.15136970281600953,
      "memory(GiB)": 118.47,
      "step": 335,
      "token_acc": 0.9577814569536424,
      "train_speed(iter/s)": 0.020508
    },
    {
      "epoch": 0.8825438027255029,
      "grad_norm": 2.515625,
      "learning_rate": 9.823313197778434e-05,
      "loss": 0.12393852472305297,
      "memory(GiB)": 118.47,
      "step": 340,
      "token_acc": 0.9630541871921182,
      "train_speed(iter/s)": 0.020525
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 5.65625,
      "learning_rate": 9.811166785176784e-05,
      "loss": 0.12096943855285644,
      "memory(GiB)": 118.47,
      "step": 345,
      "token_acc": 0.9550748752079867,
      "train_speed(iter/s)": 0.020516
    },
    {
      "epoch": 0.9085009733939,
      "grad_norm": 5.03125,
      "learning_rate": 9.798624649464323e-05,
      "loss": 0.1423720359802246,
      "memory(GiB)": 118.47,
      "step": 350,
      "token_acc": 0.9579554822753503,
      "train_speed(iter/s)": 0.020518
    },
    {
      "epoch": 0.9214795587280986,
      "grad_norm": 3.609375,
      "learning_rate": 9.785687822243812e-05,
      "loss": 0.1521735668182373,
      "memory(GiB)": 118.47,
      "step": 355,
      "token_acc": 0.9567699836867863,
      "train_speed(iter/s)": 0.020531
    },
    {
      "epoch": 0.9344581440622972,
      "grad_norm": 2.078125,
      "learning_rate": 9.772357367581779e-05,
      "loss": 0.14129271507263183,
      "memory(GiB)": 118.47,
      "step": 360,
      "token_acc": 0.9501661129568106,
      "train_speed(iter/s)": 0.020532
    },
    {
      "epoch": 0.9474367293964958,
      "grad_norm": 3.828125,
      "learning_rate": 9.758634381920981e-05,
      "loss": 0.14151201248168946,
      "memory(GiB)": 118.47,
      "step": 365,
      "token_acc": 0.9601626016260163,
      "train_speed(iter/s)": 0.020539
    },
    {
      "epoch": 0.9604153147306943,
      "grad_norm": 2.0625,
      "learning_rate": 9.744519993990232e-05,
      "loss": 0.11267189979553223,
      "memory(GiB)": 118.47,
      "step": 370,
      "token_acc": 0.9596042868920033,
      "train_speed(iter/s)": 0.020569
    },
    {
      "epoch": 0.9733939000648929,
      "grad_norm": 2.796875,
      "learning_rate": 9.730015364711557e-05,
      "loss": 0.1380028247833252,
      "memory(GiB)": 118.47,
      "step": 375,
      "token_acc": 0.9532163742690059,
      "train_speed(iter/s)": 0.02059
    },
    {
      "epoch": 0.9863724853990915,
      "grad_norm": 4.21875,
      "learning_rate": 9.715121687104715e-05,
      "loss": 0.13349549770355223,
      "memory(GiB)": 118.47,
      "step": 380,
      "token_acc": 0.9553933495539335,
      "train_speed(iter/s)": 0.020592
    },
    {
      "epoch": 0.9993510707332901,
      "grad_norm": 4.25,
      "learning_rate": 9.69984018618906e-05,
      "loss": 0.11987955570220947,
      "memory(GiB)": 118.47,
      "step": 385,
      "token_acc": 0.9604377104377104,
      "train_speed(iter/s)": 0.020608
    },
    {
      "epoch": 1.0103828682673588,
      "grad_norm": 2.890625,
      "learning_rate": 9.684172118882788e-05,
      "loss": 0.07847663760185242,
      "memory(GiB)": 118.47,
      "step": 390,
      "token_acc": 0.9701636188642926,
      "train_speed(iter/s)": 0.020645
    },
    {
      "epoch": 1.0233614536015574,
      "grad_norm": 2.578125,
      "learning_rate": 9.668118773899554e-05,
      "loss": 0.28727552890777586,
      "memory(GiB)": 118.47,
      "step": 395,
      "token_acc": 0.9520661157024793,
      "train_speed(iter/s)": 0.020648
    },
    {
      "epoch": 1.036340038935756,
      "grad_norm": 2.71875,
      "learning_rate": 9.651681471642476e-05,
      "loss": 0.10458629131317139,
      "memory(GiB)": 118.47,
      "step": 400,
      "token_acc": 0.9610927152317881,
      "train_speed(iter/s)": 0.020649
    },
    {
      "epoch": 1.0493186242699546,
      "grad_norm": 4.0625,
      "learning_rate": 9.634861564095524e-05,
      "loss": 0.07832455635070801,
      "memory(GiB)": 118.47,
      "step": 405,
      "token_acc": 0.9694041867954911,
      "train_speed(iter/s)": 0.020671
    },
    {
      "epoch": 1.0622972096041532,
      "grad_norm": 9.375,
      "learning_rate": 9.617660434712328e-05,
      "loss": 0.09629099369049073,
      "memory(GiB)": 118.47,
      "step": 410,
      "token_acc": 0.9698942229454841,
      "train_speed(iter/s)": 0.02067
    },
    {
      "epoch": 1.0752757949383518,
      "grad_norm": 1.5625,
      "learning_rate": 9.600079498302376e-05,
      "loss": 0.0883810818195343,
      "memory(GiB)": 118.47,
      "step": 415,
      "token_acc": 0.9762295081967213,
      "train_speed(iter/s)": 0.020675
    },
    {
      "epoch": 1.0882543802725504,
      "grad_norm": 2.203125,
      "learning_rate": 9.582120200914654e-05,
      "loss": 0.09818625450134277,
      "memory(GiB)": 118.47,
      "step": 420,
      "token_acc": 0.962360992301112,
      "train_speed(iter/s)": 0.020676
    },
    {
      "epoch": 1.1012329656067488,
      "grad_norm": 3.734375,
      "learning_rate": 9.563784019718703e-05,
      "loss": 0.09333943128585816,
      "memory(GiB)": 118.47,
      "step": 425,
      "token_acc": 0.9726468222043443,
      "train_speed(iter/s)": 0.020684
    },
    {
      "epoch": 1.1142115509409474,
      "grad_norm": 1.6953125,
      "learning_rate": 9.54507246288312e-05,
      "loss": 0.0676883339881897,
      "memory(GiB)": 118.47,
      "step": 430,
      "token_acc": 0.978981406628941,
      "train_speed(iter/s)": 0.020692
    },
    {
      "epoch": 1.127190136275146,
      "grad_norm": 2.390625,
      "learning_rate": 9.525987069451513e-05,
      "loss": 0.09929285049438477,
      "memory(GiB)": 118.47,
      "step": 435,
      "token_acc": 0.9760383386581469,
      "train_speed(iter/s)": 0.020704
    },
    {
      "epoch": 1.1401687216093446,
      "grad_norm": 1.7890625,
      "learning_rate": 9.506529409215907e-05,
      "loss": 0.11887519359588623,
      "memory(GiB)": 118.47,
      "step": 440,
      "token_acc": 0.956989247311828,
      "train_speed(iter/s)": 0.020705
    },
    {
      "epoch": 1.1531473069435432,
      "grad_norm": 1.6484375,
      "learning_rate": 9.486701082587634e-05,
      "loss": 0.09447284936904907,
      "memory(GiB)": 118.47,
      "step": 445,
      "token_acc": 0.9660033167495854,
      "train_speed(iter/s)": 0.020692
    },
    {
      "epoch": 1.1661258922777418,
      "grad_norm": 2.796875,
      "learning_rate": 9.466503720465692e-05,
      "loss": 0.11140995025634766,
      "memory(GiB)": 118.47,
      "step": 450,
      "token_acc": 0.9652247667514843,
      "train_speed(iter/s)": 0.020698
    },
    {
      "epoch": 1.1791044776119404,
      "grad_norm": 4.9375,
      "learning_rate": 9.445938984102606e-05,
      "loss": 0.10376272201538086,
      "memory(GiB)": 118.47,
      "step": 455,
      "token_acc": 0.9649551752241239,
      "train_speed(iter/s)": 0.020701
    },
    {
      "epoch": 1.1920830629461387,
      "grad_norm": 5.6875,
      "learning_rate": 9.42500856496779e-05,
      "loss": 0.10616805553436279,
      "memory(GiB)": 118.47,
      "step": 460,
      "token_acc": 0.9671052631578947,
      "train_speed(iter/s)": 0.020708
    },
    {
      "epoch": 1.2050616482803373,
      "grad_norm": 2.046875,
      "learning_rate": 9.403714184608411e-05,
      "loss": 0.10312038660049438,
      "memory(GiB)": 118.47,
      "step": 465,
      "token_acc": 0.9700404858299595,
      "train_speed(iter/s)": 0.02071
    },
    {
      "epoch": 1.218040233614536,
      "grad_norm": 1.6796875,
      "learning_rate": 9.382057594507805e-05,
      "loss": 0.11268962621688842,
      "memory(GiB)": 118.47,
      "step": 470,
      "token_acc": 0.9643146796431468,
      "train_speed(iter/s)": 0.020713
    },
    {
      "epoch": 1.2310188189487346,
      "grad_norm": 2.125,
      "learning_rate": 9.360040575941405e-05,
      "loss": 0.11029822826385498,
      "memory(GiB)": 118.47,
      "step": 475,
      "token_acc": 0.961439588688946,
      "train_speed(iter/s)": 0.02071
    },
    {
      "epoch": 1.2439974042829332,
      "grad_norm": 2.734375,
      "learning_rate": 9.337664939830234e-05,
      "loss": 0.08803771138191223,
      "memory(GiB)": 118.47,
      "step": 480,
      "token_acc": 0.9737919737919738,
      "train_speed(iter/s)": 0.020707
    },
    {
      "epoch": 1.2569759896171318,
      "grad_norm": 2.46875,
      "learning_rate": 9.314932526591956e-05,
      "loss": 0.12014460563659668,
      "memory(GiB)": 118.47,
      "step": 485,
      "token_acc": 0.9600665557404326,
      "train_speed(iter/s)": 0.020714
    },
    {
      "epoch": 1.2699545749513304,
      "grad_norm": 1.2890625,
      "learning_rate": 9.291845205989495e-05,
      "loss": 0.061967766284942626,
      "memory(GiB)": 118.47,
      "step": 490,
      "token_acc": 0.9819227608874281,
      "train_speed(iter/s)": 0.020719
    },
    {
      "epoch": 1.282933160285529,
      "grad_norm": 3.328125,
      "learning_rate": 9.26840487697725e-05,
      "loss": 0.0871744453907013,
      "memory(GiB)": 118.47,
      "step": 495,
      "token_acc": 0.971240755957272,
      "train_speed(iter/s)": 0.020714
    },
    {
      "epoch": 1.2959117456197276,
      "grad_norm": 3.53125,
      "learning_rate": 9.244613467544903e-05,
      "loss": 0.11612107753753662,
      "memory(GiB)": 118.47,
      "step": 500,
      "token_acc": 0.9636062861869313,
      "train_speed(iter/s)": 0.020724
    },
    {
      "epoch": 1.308890330953926,
      "grad_norm": 2.453125,
      "learning_rate": 9.220472934558837e-05,
      "loss": 0.09390705227851867,
      "memory(GiB)": 118.47,
      "step": 505,
      "token_acc": 0.9699499165275459,
      "train_speed(iter/s)": 0.020673
    },
    {
      "epoch": 1.3218689162881245,
      "grad_norm": 3.546875,
      "learning_rate": 9.195985263601186e-05,
      "loss": 0.09252657294273377,
      "memory(GiB)": 118.47,
      "step": 510,
      "token_acc": 0.9639865996649917,
      "train_speed(iter/s)": 0.020679
    },
    {
      "epoch": 1.3348475016223231,
      "grad_norm": 1.4765625,
      "learning_rate": 9.171152468806516e-05,
      "loss": 0.08763872385025025,
      "memory(GiB)": 118.47,
      "step": 515,
      "token_acc": 0.9736408566721582,
      "train_speed(iter/s)": 0.020688
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 1.9609375,
      "learning_rate": 9.145976592696163e-05,
      "loss": 0.08335199356079101,
      "memory(GiB)": 118.47,
      "step": 520,
      "token_acc": 0.9696472518457753,
      "train_speed(iter/s)": 0.020699
    },
    {
      "epoch": 1.3608046722907203,
      "grad_norm": 1.734375,
      "learning_rate": 9.120459706010233e-05,
      "loss": 0.08465138673782349,
      "memory(GiB)": 118.47,
      "step": 525,
      "token_acc": 0.9754028837998303,
      "train_speed(iter/s)": 0.020692
    },
    {
      "epoch": 1.373783257624919,
      "grad_norm": 2.21875,
      "learning_rate": 9.094603907537276e-05,
      "loss": 0.08926347494125367,
      "memory(GiB)": 118.47,
      "step": 530,
      "token_acc": 0.9690893901420217,
      "train_speed(iter/s)": 0.020699
    },
    {
      "epoch": 1.3867618429591175,
      "grad_norm": 2.28125,
      "learning_rate": 9.06841132394167e-05,
      "loss": 0.07581965327262878,
      "memory(GiB)": 118.47,
      "step": 535,
      "token_acc": 0.9726141078838174,
      "train_speed(iter/s)": 0.020705
    },
    {
      "epoch": 1.399740428293316,
      "grad_norm": 1.5,
      "learning_rate": 9.04188410958869e-05,
      "loss": 0.08940108418464661,
      "memory(GiB)": 118.47,
      "step": 540,
      "token_acc": 0.9730639730639731,
      "train_speed(iter/s)": 0.020709
    },
    {
      "epoch": 1.4127190136275147,
      "grad_norm": 3.90625,
      "learning_rate": 9.015024446367315e-05,
      "loss": 0.10531409978866577,
      "memory(GiB)": 118.47,
      "step": 545,
      "token_acc": 0.9656862745098039,
      "train_speed(iter/s)": 0.020718
    },
    {
      "epoch": 1.425697598961713,
      "grad_norm": 1.09375,
      "learning_rate": 8.987834543510765e-05,
      "loss": 0.10772655010223389,
      "memory(GiB)": 118.47,
      "step": 550,
      "token_acc": 0.9670608108108109,
      "train_speed(iter/s)": 0.020719
    },
    {
      "epoch": 1.4386761842959117,
      "grad_norm": 0.9921875,
      "learning_rate": 8.960316637414787e-05,
      "loss": 0.09822122454643249,
      "memory(GiB)": 118.47,
      "step": 555,
      "token_acc": 0.9698205546492659,
      "train_speed(iter/s)": 0.020733
    },
    {
      "epoch": 1.4516547696301103,
      "grad_norm": 3.953125,
      "learning_rate": 8.932472991453712e-05,
      "loss": 0.08714378476142884,
      "memory(GiB)": 118.47,
      "step": 560,
      "token_acc": 0.966988727858293,
      "train_speed(iter/s)": 0.020744
    },
    {
      "epoch": 1.464633354964309,
      "grad_norm": 3.6875,
      "learning_rate": 8.904305895794292e-05,
      "loss": 0.1327031970024109,
      "memory(GiB)": 118.47,
      "step": 565,
      "token_acc": 0.9598689598689598,
      "train_speed(iter/s)": 0.020746
    },
    {
      "epoch": 1.4776119402985075,
      "grad_norm": 2.4375,
      "learning_rate": 8.87581766720732e-05,
      "loss": 0.08868688941001893,
      "memory(GiB)": 118.47,
      "step": 570,
      "token_acc": 0.9679539852095317,
      "train_speed(iter/s)": 0.020754
    },
    {
      "epoch": 1.490590525632706,
      "grad_norm": 2.296875,
      "learning_rate": 8.847010648877094e-05,
      "loss": 0.07921120524406433,
      "memory(GiB)": 118.47,
      "step": 575,
      "token_acc": 0.9747145187601958,
      "train_speed(iter/s)": 0.020753
    },
    {
      "epoch": 1.5035691109669047,
      "grad_norm": 0.96875,
      "learning_rate": 8.817887210208668e-05,
      "loss": 0.0657312035560608,
      "memory(GiB)": 118.47,
      "step": 580,
      "token_acc": 0.9798387096774194,
      "train_speed(iter/s)": 0.020759
    },
    {
      "epoch": 1.516547696301103,
      "grad_norm": 1.2265625,
      "learning_rate": 8.788449746632977e-05,
      "loss": 0.06689350008964538,
      "memory(GiB)": 118.47,
      "step": 585,
      "token_acc": 0.974485596707819,
      "train_speed(iter/s)": 0.020758
    },
    {
      "epoch": 1.529526281635302,
      "grad_norm": 2.03125,
      "learning_rate": 8.758700679409804e-05,
      "loss": 0.05711461901664734,
      "memory(GiB)": 118.47,
      "step": 590,
      "token_acc": 0.9784860557768924,
      "train_speed(iter/s)": 0.02076
    },
    {
      "epoch": 1.5425048669695003,
      "grad_norm": 1.546875,
      "learning_rate": 8.72864245542864e-05,
      "loss": 0.09383676052093506,
      "memory(GiB)": 118.47,
      "step": 595,
      "token_acc": 0.9712595097210481,
      "train_speed(iter/s)": 0.020756
    },
    {
      "epoch": 1.5554834523036989,
      "grad_norm": 1.140625,
      "learning_rate": 8.69827754700741e-05,
      "loss": 0.08457305431365966,
      "memory(GiB)": 118.47,
      "step": 600,
      "token_acc": 0.9700404858299595,
      "train_speed(iter/s)": 0.020755
    },
    {
      "epoch": 1.5684620376378975,
      "grad_norm": 1.171875,
      "learning_rate": 8.667608451689134e-05,
      "loss": 0.05290307998657227,
      "memory(GiB)": 118.47,
      "step": 605,
      "token_acc": 0.9830097087378641,
      "train_speed(iter/s)": 0.020756
    },
    {
      "epoch": 1.581440622972096,
      "grad_norm": 3.734375,
      "learning_rate": 8.6366376920365e-05,
      "loss": 0.0879642128944397,
      "memory(GiB)": 118.47,
      "step": 610,
      "token_acc": 0.9723577235772358,
      "train_speed(iter/s)": 0.020761
    },
    {
      "epoch": 1.5944192083062947,
      "grad_norm": 1.1875,
      "learning_rate": 8.605367815424372e-05,
      "loss": 0.05958070755004883,
      "memory(GiB)": 118.47,
      "step": 615,
      "token_acc": 0.9812244897959184,
      "train_speed(iter/s)": 0.020757
    },
    {
      "epoch": 1.607397793640493,
      "grad_norm": 14.3125,
      "learning_rate": 8.573801393830281e-05,
      "loss": 0.15476706027984619,
      "memory(GiB)": 118.47,
      "step": 620,
      "token_acc": 0.9589958158995816,
      "train_speed(iter/s)": 0.020758
    },
    {
      "epoch": 1.6203763789746919,
      "grad_norm": 1.484375,
      "learning_rate": 8.54194102362286e-05,
      "loss": 0.08144987821578979,
      "memory(GiB)": 118.47,
      "step": 625,
      "token_acc": 0.977796052631579,
      "train_speed(iter/s)": 0.020753
    },
    {
      "epoch": 1.6333549643088903,
      "grad_norm": 2.375,
      "learning_rate": 8.509789325348306e-05,
      "loss": 0.12072479724884033,
      "memory(GiB)": 118.47,
      "step": 630,
      "token_acc": 0.9649122807017544,
      "train_speed(iter/s)": 0.020756
    },
    {
      "epoch": 1.6463335496430889,
      "grad_norm": 1.171875,
      "learning_rate": 8.477348943514828e-05,
      "loss": 0.057410746812820435,
      "memory(GiB)": 118.47,
      "step": 635,
      "token_acc": 0.9805510534846029,
      "train_speed(iter/s)": 0.020755
    },
    {
      "epoch": 1.6593121349772875,
      "grad_norm": 2.890625,
      "learning_rate": 8.444622546375136e-05,
      "loss": 0.08364247083663941,
      "memory(GiB)": 118.47,
      "step": 640,
      "token_acc": 0.9754500818330606,
      "train_speed(iter/s)": 0.020763
    },
    {
      "epoch": 1.672290720311486,
      "grad_norm": 1.4296875,
      "learning_rate": 8.411612825706977e-05,
      "loss": 0.0733809769153595,
      "memory(GiB)": 118.47,
      "step": 645,
      "token_acc": 0.9763458401305057,
      "train_speed(iter/s)": 0.020764
    },
    {
      "epoch": 1.6852693056456847,
      "grad_norm": 2.375,
      "learning_rate": 8.378322496591727e-05,
      "loss": 0.0672804057598114,
      "memory(GiB)": 118.47,
      "step": 650,
      "token_acc": 0.9800498753117207,
      "train_speed(iter/s)": 0.020765
    },
    {
      "epoch": 1.698247890979883,
      "grad_norm": 3.328125,
      "learning_rate": 8.34475429719108e-05,
      "loss": 0.09220197200775146,
      "memory(GiB)": 118.47,
      "step": 655,
      "token_acc": 0.971900826446281,
      "train_speed(iter/s)": 0.020763
    },
    {
      "epoch": 1.7112264763140819,
      "grad_norm": 2.109375,
      "learning_rate": 8.310910988521836e-05,
      "loss": 0.09654330611228942,
      "memory(GiB)": 118.47,
      "step": 660,
      "token_acc": 0.9690301548492257,
      "train_speed(iter/s)": 0.020766
    },
    {
      "epoch": 1.7242050616482802,
      "grad_norm": 1.609375,
      "learning_rate": 8.276795354228786e-05,
      "loss": 0.09199822545051575,
      "memory(GiB)": 118.47,
      "step": 665,
      "token_acc": 0.9744027303754266,
      "train_speed(iter/s)": 0.020764
    },
    {
      "epoch": 1.737183646982479,
      "grad_norm": 1.9609375,
      "learning_rate": 8.242410200355772e-05,
      "loss": 0.08355655074119568,
      "memory(GiB)": 118.47,
      "step": 670,
      "token_acc": 0.9717607973421927,
      "train_speed(iter/s)": 0.020765
    },
    {
      "epoch": 1.7501622323166774,
      "grad_norm": 1.3671875,
      "learning_rate": 8.207758355114883e-05,
      "loss": 0.07680343985557556,
      "memory(GiB)": 118.47,
      "step": 675,
      "token_acc": 0.9784768211920529,
      "train_speed(iter/s)": 0.020769
    },
    {
      "epoch": 1.763140817650876,
      "grad_norm": 1.1328125,
      "learning_rate": 8.172842668653829e-05,
      "loss": 0.05557375550270081,
      "memory(GiB)": 118.47,
      "step": 680,
      "token_acc": 0.983150800336984,
      "train_speed(iter/s)": 0.02077
    },
    {
      "epoch": 1.7761194029850746,
      "grad_norm": 3.46875,
      "learning_rate": 8.137666012821514e-05,
      "loss": 0.0882513701915741,
      "memory(GiB)": 118.47,
      "step": 685,
      "token_acc": 0.9752770673486786,
      "train_speed(iter/s)": 0.020769
    },
    {
      "epoch": 1.7890979883192732,
      "grad_norm": 2.96875,
      "learning_rate": 8.102231280931825e-05,
      "loss": 0.10723648071289063,
      "memory(GiB)": 118.47,
      "step": 690,
      "token_acc": 0.9713584288052373,
      "train_speed(iter/s)": 0.020778
    },
    {
      "epoch": 1.8020765736534718,
      "grad_norm": 1.5078125,
      "learning_rate": 8.066541387525656e-05,
      "loss": 0.07498295307159424,
      "memory(GiB)": 118.47,
      "step": 695,
      "token_acc": 0.9777043765483072,
      "train_speed(iter/s)": 0.020784
    },
    {
      "epoch": 1.8150551589876702,
      "grad_norm": 2.84375,
      "learning_rate": 8.030599268131178e-05,
      "loss": 0.06255126595497132,
      "memory(GiB)": 118.47,
      "step": 700,
      "token_acc": 0.9788617886178862,
      "train_speed(iter/s)": 0.020786
    },
    {
      "epoch": 1.828033744321869,
      "grad_norm": 2.65625,
      "learning_rate": 7.994407879022396e-05,
      "loss": 0.09357861876487732,
      "memory(GiB)": 118.47,
      "step": 705,
      "token_acc": 0.9719703215169002,
      "train_speed(iter/s)": 0.02079
    },
    {
      "epoch": 1.8410123296560674,
      "grad_norm": 2.078125,
      "learning_rate": 7.957970196975989e-05,
      "loss": 0.04933435618877411,
      "memory(GiB)": 118.47,
      "step": 710,
      "token_acc": 0.984375,
      "train_speed(iter/s)": 0.020784
    },
    {
      "epoch": 1.853990914990266,
      "grad_norm": 1.2421875,
      "learning_rate": 7.921289219026465e-05,
      "loss": 0.04773304164409638,
      "memory(GiB)": 118.47,
      "step": 715,
      "token_acc": 0.9835390946502057,
      "train_speed(iter/s)": 0.020777
    },
    {
      "epoch": 1.8669695003244646,
      "grad_norm": 2.40625,
      "learning_rate": 7.884367962219659e-05,
      "loss": 0.10059411525726318,
      "memory(GiB)": 118.47,
      "step": 720,
      "token_acc": 0.9690117252931323,
      "train_speed(iter/s)": 0.020775
    },
    {
      "epoch": 1.8799480856586632,
      "grad_norm": 2.65625,
      "learning_rate": 7.847209463364573e-05,
      "loss": 0.0753164291381836,
      "memory(GiB)": 118.47,
      "step": 725,
      "token_acc": 0.9772727272727273,
      "train_speed(iter/s)": 0.020777
    },
    {
      "epoch": 1.8929266709928618,
      "grad_norm": 3.03125,
      "learning_rate": 7.809816778783594e-05,
      "loss": 0.06520425677299499,
      "memory(GiB)": 118.47,
      "step": 730,
      "token_acc": 0.9786359901396877,
      "train_speed(iter/s)": 0.020783
    },
    {
      "epoch": 1.9059052563270602,
      "grad_norm": 1.9609375,
      "learning_rate": 7.772192984061113e-05,
      "loss": 0.08142446279525757,
      "memory(GiB)": 118.47,
      "step": 735,
      "token_acc": 0.9693454846727423,
      "train_speed(iter/s)": 0.020794
    },
    {
      "epoch": 1.918883841661259,
      "grad_norm": 1.03125,
      "learning_rate": 7.734341173790551e-05,
      "loss": 0.07234652638435364,
      "memory(GiB)": 118.47,
      "step": 740,
      "token_acc": 0.9788617886178862,
      "train_speed(iter/s)": 0.020798
    },
    {
      "epoch": 1.9318624269954574,
      "grad_norm": 2.03125,
      "learning_rate": 7.696264461319831e-05,
      "loss": 0.07108655571937561,
      "memory(GiB)": 118.47,
      "step": 745,
      "token_acc": 0.975103734439834,
      "train_speed(iter/s)": 0.020799
    },
    {
      "epoch": 1.9448410123296562,
      "grad_norm": 2.125,
      "learning_rate": 7.657965978495299e-05,
      "loss": 0.06093553900718689,
      "memory(GiB)": 118.47,
      "step": 750,
      "token_acc": 0.9778325123152709,
      "train_speed(iter/s)": 0.020808
    },
    {
      "epoch": 1.9578195976638546,
      "grad_norm": 1.28125,
      "learning_rate": 7.619448875404121e-05,
      "loss": 0.06928490400314331,
      "memory(GiB)": 118.47,
      "step": 755,
      "token_acc": 0.9735973597359736,
      "train_speed(iter/s)": 0.020817
    },
    {
      "epoch": 1.9707981829980532,
      "grad_norm": 1.7734375,
      "learning_rate": 7.580716320115195e-05,
      "loss": 0.07247484922409057,
      "memory(GiB)": 118.47,
      "step": 760,
      "token_acc": 0.9713322091062394,
      "train_speed(iter/s)": 0.020819
    },
    {
      "epoch": 1.9837767683322518,
      "grad_norm": 0.9375,
      "learning_rate": 7.541771498418575e-05,
      "loss": 0.07819901704788208,
      "memory(GiB)": 118.47,
      "step": 765,
      "token_acc": 0.9759136212624585,
      "train_speed(iter/s)": 0.02082
    },
    {
      "epoch": 1.9967553536664504,
      "grad_norm": 1.5859375,
      "learning_rate": 7.502617613563423e-05,
      "loss": 0.06084737181663513,
      "memory(GiB)": 118.47,
      "step": 770,
      "token_acc": 0.978369384359401,
      "train_speed(iter/s)": 0.020823
    },
    {
      "epoch": 2.007787151200519,
      "grad_norm": 1.625,
      "learning_rate": 7.463257885994552e-05,
      "loss": 0.04811320304870605,
      "memory(GiB)": 118.47,
      "step": 775,
      "token_acc": 0.980980980980981,
      "train_speed(iter/s)": 0.020841
    },
    {
      "epoch": 2.0207657365347176,
      "grad_norm": 0.427734375,
      "learning_rate": 7.423695553087537e-05,
      "loss": 0.045494526624679565,
      "memory(GiB)": 118.47,
      "step": 780,
      "token_acc": 0.985233798195242,
      "train_speed(iter/s)": 0.020848
    },
    {
      "epoch": 2.0337443218689164,
      "grad_norm": 5.78125,
      "learning_rate": 7.383933868882438e-05,
      "loss": 0.05818610191345215,
      "memory(GiB)": 118.47,
      "step": 785,
      "token_acc": 0.9794168096054888,
      "train_speed(iter/s)": 0.020848
    },
    {
      "epoch": 2.046722907203115,
      "grad_norm": 0.890625,
      "learning_rate": 7.343976103816148e-05,
      "loss": 0.037669533491134645,
      "memory(GiB)": 118.47,
      "step": 790,
      "token_acc": 0.9900083263946711,
      "train_speed(iter/s)": 0.020846
    },
    {
      "epoch": 2.0597014925373136,
      "grad_norm": 2.9375,
      "learning_rate": 7.303825544453404e-05,
      "loss": 0.09385978579521179,
      "memory(GiB)": 118.47,
      "step": 795,
      "token_acc": 0.9735973597359736,
      "train_speed(iter/s)": 0.020851
    },
    {
      "epoch": 2.072680077871512,
      "grad_norm": 1.3359375,
      "learning_rate": 7.263485493216458e-05,
      "loss": 0.05340880751609802,
      "memory(GiB)": 118.47,
      "step": 800,
      "token_acc": 0.985353946297803,
      "train_speed(iter/s)": 0.020854
    },
    {
      "epoch": 2.0856586632057104,
      "grad_norm": 1.0625,
      "learning_rate": 7.222959268113452e-05,
      "loss": 0.05104266405105591,
      "memory(GiB)": 118.47,
      "step": 805,
      "token_acc": 0.9834574028122415,
      "train_speed(iter/s)": 0.020863
    },
    {
      "epoch": 2.098637248539909,
      "grad_norm": 1.828125,
      "learning_rate": 7.182250202465509e-05,
      "loss": 0.052119660377502444,
      "memory(GiB)": 118.47,
      "step": 810,
      "token_acc": 0.9854721549636803,
      "train_speed(iter/s)": 0.020866
    },
    {
      "epoch": 2.1116158338741076,
      "grad_norm": 1.96875,
      "learning_rate": 7.14136164463256e-05,
      "loss": 0.06113920211791992,
      "memory(GiB)": 118.47,
      "step": 815,
      "token_acc": 0.9816360601001669,
      "train_speed(iter/s)": 0.02087
    },
    {
      "epoch": 2.1245944192083064,
      "grad_norm": 1.3125,
      "learning_rate": 7.100296957737948e-05,
      "loss": 0.06530060768127441,
      "memory(GiB)": 118.47,
      "step": 820,
      "token_acc": 0.9808811305070657,
      "train_speed(iter/s)": 0.020872
    },
    {
      "epoch": 2.1375730045425048,
      "grad_norm": 2.0,
      "learning_rate": 7.059059519391794e-05,
      "loss": 0.07137447595596313,
      "memory(GiB)": 118.47,
      "step": 825,
      "token_acc": 0.9771801140994295,
      "train_speed(iter/s)": 0.020874
    },
    {
      "epoch": 2.1505515898767036,
      "grad_norm": 1.4609375,
      "learning_rate": 7.017652721413196e-05,
      "loss": 0.045847800374031064,
      "memory(GiB)": 118.47,
      "step": 830,
      "token_acc": 0.9824267782426779,
      "train_speed(iter/s)": 0.020875
    },
    {
      "epoch": 2.163530175210902,
      "grad_norm": 1.2265625,
      "learning_rate": 6.976079969551246e-05,
      "loss": 0.04792321026325226,
      "memory(GiB)": 118.47,
      "step": 835,
      "token_acc": 0.9875415282392026,
      "train_speed(iter/s)": 0.020875
    },
    {
      "epoch": 2.176508760545101,
      "grad_norm": 1.125,
      "learning_rate": 6.934344683204901e-05,
      "loss": 0.04856709241867065,
      "memory(GiB)": 118.47,
      "step": 840,
      "token_acc": 0.9852459016393442,
      "train_speed(iter/s)": 0.020879
    },
    {
      "epoch": 2.189487345879299,
      "grad_norm": 1.4453125,
      "learning_rate": 6.892450295141736e-05,
      "loss": 0.05035427212715149,
      "memory(GiB)": 118.47,
      "step": 845,
      "token_acc": 0.9832635983263598,
      "train_speed(iter/s)": 0.020881
    },
    {
      "epoch": 2.2024659312134975,
      "grad_norm": 1.5703125,
      "learning_rate": 6.8504002512156e-05,
      "loss": 0.03542698323726654,
      "memory(GiB)": 118.47,
      "step": 850,
      "token_acc": 0.9884488448844885,
      "train_speed(iter/s)": 0.02088
    },
    {
      "epoch": 2.2154445165476964,
      "grad_norm": 1.8515625,
      "learning_rate": 6.808198010083185e-05,
      "loss": 0.03674182891845703,
      "memory(GiB)": 118.47,
      "step": 855,
      "token_acc": 0.9859271523178808,
      "train_speed(iter/s)": 0.020878
    },
    {
      "epoch": 2.2284231018818947,
      "grad_norm": 0.408203125,
      "learning_rate": 6.765847042919542e-05,
      "loss": 0.0341266006231308,
      "memory(GiB)": 118.47,
      "step": 860,
      "token_acc": 0.9859504132231405,
      "train_speed(iter/s)": 0.020878
    },
    {
      "epoch": 2.2414016872160936,
      "grad_norm": 0.8828125,
      "learning_rate": 6.723350833132595e-05,
      "loss": 0.047522234916687014,
      "memory(GiB)": 118.47,
      "step": 865,
      "token_acc": 0.9858333333333333,
      "train_speed(iter/s)": 0.020875
    },
    {
      "epoch": 2.254380272550292,
      "grad_norm": 0.6953125,
      "learning_rate": 6.680712876076605e-05,
      "loss": 0.04341578781604767,
      "memory(GiB)": 118.47,
      "step": 870,
      "token_acc": 0.9887005649717514,
      "train_speed(iter/s)": 0.020876
    },
    {
      "epoch": 2.2673588578844908,
      "grad_norm": 1.25,
      "learning_rate": 6.637936678764684e-05,
      "loss": 0.04864707887172699,
      "memory(GiB)": 118.47,
      "step": 875,
      "token_acc": 0.9837000814995925,
      "train_speed(iter/s)": 0.020885
    },
    {
      "epoch": 2.280337443218689,
      "grad_norm": 0.7421875,
      "learning_rate": 6.595025759580342e-05,
      "loss": 0.02881385087966919,
      "memory(GiB)": 118.47,
      "step": 880,
      "token_acc": 0.9928684627575277,
      "train_speed(iter/s)": 0.02089
    },
    {
      "epoch": 2.293316028552888,
      "grad_norm": 1.0625,
      "learning_rate": 6.551983647988089e-05,
      "loss": 0.03752845525741577,
      "memory(GiB)": 118.47,
      "step": 885,
      "token_acc": 0.9892650701899257,
      "train_speed(iter/s)": 0.020893
    },
    {
      "epoch": 2.3062946138870863,
      "grad_norm": 0.70703125,
      "learning_rate": 6.508813884243138e-05,
      "loss": 0.04644352495670319,
      "memory(GiB)": 118.47,
      "step": 890,
      "token_acc": 0.9877149877149877,
      "train_speed(iter/s)": 0.020898
    },
    {
      "epoch": 2.3192731992212847,
      "grad_norm": 2.21875,
      "learning_rate": 6.465520019100217e-05,
      "loss": 0.04913321435451508,
      "memory(GiB)": 118.47,
      "step": 895,
      "token_acc": 0.9795221843003413,
      "train_speed(iter/s)": 0.020898
    },
    {
      "epoch": 2.3322517845554835,
      "grad_norm": 1.59375,
      "learning_rate": 6.422105613521512e-05,
      "loss": 0.03812339305877686,
      "memory(GiB)": 118.47,
      "step": 900,
      "token_acc": 0.986938775510204,
      "train_speed(iter/s)": 0.020893
    },
    {
      "epoch": 2.345230369889682,
      "grad_norm": 1.5625,
      "learning_rate": 6.378574238383777e-05,
      "loss": 0.04143716990947723,
      "memory(GiB)": 118.47,
      "step": 905,
      "token_acc": 0.9850746268656716,
      "train_speed(iter/s)": 0.020891
    },
    {
      "epoch": 2.3582089552238807,
      "grad_norm": 1.6875,
      "learning_rate": 6.334929474184621e-05,
      "loss": 0.05560382604598999,
      "memory(GiB)": 118.47,
      "step": 910,
      "token_acc": 0.9827018121911038,
      "train_speed(iter/s)": 0.020899
    },
    {
      "epoch": 2.371187540558079,
      "grad_norm": 1.7890625,
      "learning_rate": 6.29117491074802e-05,
      "loss": 0.044425755739212036,
      "memory(GiB)": 118.47,
      "step": 915,
      "token_acc": 0.9833610648918469,
      "train_speed(iter/s)": 0.020901
    },
    {
      "epoch": 2.3841661258922775,
      "grad_norm": 0.8359375,
      "learning_rate": 6.247314146929038e-05,
      "loss": 0.050447821617126465,
      "memory(GiB)": 118.47,
      "step": 920,
      "token_acc": 0.9852700490998363,
      "train_speed(iter/s)": 0.020905
    },
    {
      "epoch": 2.3971447112264763,
      "grad_norm": 0.5,
      "learning_rate": 6.203350790317825e-05,
      "loss": 0.022633786499500274,
      "memory(GiB)": 118.47,
      "step": 925,
      "token_acc": 0.9933388842631141,
      "train_speed(iter/s)": 0.020904
    },
    {
      "epoch": 2.4101232965606747,
      "grad_norm": 1.2109375,
      "learning_rate": 6.159288456942884e-05,
      "loss": 0.04340067207813263,
      "memory(GiB)": 118.47,
      "step": 930,
      "token_acc": 0.9845024469820555,
      "train_speed(iter/s)": 0.020906
    },
    {
      "epoch": 2.4231018818948735,
      "grad_norm": 1.2890625,
      "learning_rate": 6.115130770973658e-05,
      "loss": 0.055285471677780154,
      "memory(GiB)": 118.47,
      "step": 935,
      "token_acc": 0.9786008230452675,
      "train_speed(iter/s)": 0.020909
    },
    {
      "epoch": 2.436080467229072,
      "grad_norm": 1.25,
      "learning_rate": 6.0708813644224304e-05,
      "loss": 0.04441152215003967,
      "memory(GiB)": 118.47,
      "step": 940,
      "token_acc": 0.9883040935672515,
      "train_speed(iter/s)": 0.020911
    },
    {
      "epoch": 2.4490590525632707,
      "grad_norm": 1.6015625,
      "learning_rate": 6.026543876845586e-05,
      "loss": 0.05162730813026428,
      "memory(GiB)": 118.47,
      "step": 945,
      "token_acc": 0.9811629811629812,
      "train_speed(iter/s)": 0.020914
    },
    {
      "epoch": 2.462037637897469,
      "grad_norm": 1.0546875,
      "learning_rate": 5.9821219550442654e-05,
      "loss": 0.055637669563293454,
      "memory(GiB)": 118.47,
      "step": 950,
      "token_acc": 0.9836199836199836,
      "train_speed(iter/s)": 0.020911
    },
    {
      "epoch": 2.475016223231668,
      "grad_norm": 5.96875,
      "learning_rate": 5.937619252764405e-05,
      "loss": 0.0420777291059494,
      "memory(GiB)": 118.47,
      "step": 955,
      "token_acc": 0.9869494290375204,
      "train_speed(iter/s)": 0.020911
    },
    {
      "epoch": 2.4879948085658663,
      "grad_norm": 1.1875,
      "learning_rate": 5.89303943039621e-05,
      "loss": 0.04367494285106659,
      "memory(GiB)": 118.47,
      "step": 960,
      "token_acc": 0.9859154929577465,
      "train_speed(iter/s)": 0.020917
    },
    {
      "epoch": 2.5009733939000647,
      "grad_norm": 0.416015625,
      "learning_rate": 5.848386154673091e-05,
      "loss": 0.03488150537014008,
      "memory(GiB)": 118.47,
      "step": 965,
      "token_acc": 0.9899328859060402,
      "train_speed(iter/s)": 0.02092
    },
    {
      "epoch": 2.5139519792342635,
      "grad_norm": 0.73828125,
      "learning_rate": 5.8036630983700644e-05,
      "loss": 0.023914022743701933,
      "memory(GiB)": 118.47,
      "step": 970,
      "token_acc": 0.9895414320193081,
      "train_speed(iter/s)": 0.02092
    },
    {
      "epoch": 2.526930564568462,
      "grad_norm": 1.359375,
      "learning_rate": 5.7588739400016676e-05,
      "loss": 0.05280606150627136,
      "memory(GiB)": 118.47,
      "step": 975,
      "token_acc": 0.9850374064837906,
      "train_speed(iter/s)": 0.020918
    },
    {
      "epoch": 2.5399091499026607,
      "grad_norm": 1.1796875,
      "learning_rate": 5.7140223635194034e-05,
      "loss": 0.04982930719852448,
      "memory(GiB)": 118.47,
      "step": 980,
      "token_acc": 0.9808333333333333,
      "train_speed(iter/s)": 0.020914
    },
    {
      "epoch": 2.552887735236859,
      "grad_norm": 1.234375,
      "learning_rate": 5.669112058008712e-05,
      "loss": 0.03910036683082581,
      "memory(GiB)": 118.47,
      "step": 985,
      "token_acc": 0.9869067103109657,
      "train_speed(iter/s)": 0.02092
    },
    {
      "epoch": 2.565866320571058,
      "grad_norm": 0.5234375,
      "learning_rate": 5.624146717385563e-05,
      "loss": 0.0397420197725296,
      "memory(GiB)": 118.47,
      "step": 990,
      "token_acc": 0.9874055415617129,
      "train_speed(iter/s)": 0.020921
    },
    {
      "epoch": 2.5788449059052563,
      "grad_norm": 1.625,
      "learning_rate": 5.579130040092613e-05,
      "loss": 0.04704872071743012,
      "memory(GiB)": 118.47,
      "step": 995,
      "token_acc": 0.9854486661277284,
      "train_speed(iter/s)": 0.020919
    },
    {
      "epoch": 2.591823491239455,
      "grad_norm": 0.640625,
      "learning_rate": 5.53406572879501e-05,
      "loss": 0.025661391019821168,
      "memory(GiB)": 118.47,
      "step": 1000,
      "token_acc": 0.9893179950698439,
      "train_speed(iter/s)": 0.020923
    },
    {
      "epoch": 2.6048020765736535,
      "grad_norm": 1.453125,
      "learning_rate": 5.488957490075846e-05,
      "loss": 0.04738643765449524,
      "memory(GiB)": 118.47,
      "step": 1005,
      "token_acc": 0.9881756756756757,
      "train_speed(iter/s)": 0.020894
    },
    {
      "epoch": 2.617780661907852,
      "grad_norm": 1.7421875,
      "learning_rate": 5.443809034131283e-05,
      "loss": 0.033944371342659,
      "memory(GiB)": 118.47,
      "step": 1010,
      "token_acc": 0.9890664423885618,
      "train_speed(iter/s)": 0.020898
    },
    {
      "epoch": 2.6307592472420507,
      "grad_norm": 2.09375,
      "learning_rate": 5.3986240744653916e-05,
      "loss": 0.049440020322799684,
      "memory(GiB)": 118.47,
      "step": 1015,
      "token_acc": 0.9846938775510204,
      "train_speed(iter/s)": 0.020897
    },
    {
      "epoch": 2.643737832576249,
      "grad_norm": 0.8046875,
      "learning_rate": 5.3534063275847065e-05,
      "loss": 0.03849257528781891,
      "memory(GiB)": 118.47,
      "step": 1020,
      "token_acc": 0.9902439024390244,
      "train_speed(iter/s)": 0.0209
    },
    {
      "epoch": 2.656716417910448,
      "grad_norm": 0.734375,
      "learning_rate": 5.3081595126925444e-05,
      "loss": 0.026245176792144775,
      "memory(GiB)": 118.47,
      "step": 1025,
      "token_acc": 0.9879614767255217,
      "train_speed(iter/s)": 0.0209
    },
    {
      "epoch": 2.6696950032446463,
      "grad_norm": 1.2421875,
      "learning_rate": 5.2628873513830914e-05,
      "loss": 0.031640535593032836,
      "memory(GiB)": 118.47,
      "step": 1030,
      "token_acc": 0.9879227053140096,
      "train_speed(iter/s)": 0.020902
    },
    {
      "epoch": 2.6826735885788446,
      "grad_norm": 1.09375,
      "learning_rate": 5.2175935673353035e-05,
      "loss": 0.0378518670797348,
      "memory(GiB)": 118.47,
      "step": 1035,
      "token_acc": 0.9883040935672515,
      "train_speed(iter/s)": 0.020904
    },
    {
      "epoch": 2.6956521739130435,
      "grad_norm": 1.1328125,
      "learning_rate": 5.1722818860066245e-05,
      "loss": 0.022336561977863312,
      "memory(GiB)": 118.47,
      "step": 1040,
      "token_acc": 0.9894991922455574,
      "train_speed(iter/s)": 0.020904
    },
    {
      "epoch": 2.7086307592472423,
      "grad_norm": 0.8671875,
      "learning_rate": 5.126956034326573e-05,
      "loss": 0.02416083961725235,
      "memory(GiB)": 118.47,
      "step": 1045,
      "token_acc": 0.9918300653594772,
      "train_speed(iter/s)": 0.020901
    },
    {
      "epoch": 2.7216093445814407,
      "grad_norm": 1.3125,
      "learning_rate": 5.0816197403901876e-05,
      "loss": 0.03095039427280426,
      "memory(GiB)": 118.47,
      "step": 1050,
      "token_acc": 0.9918500407497962,
      "train_speed(iter/s)": 0.020906
    },
    {
      "epoch": 2.734587929915639,
      "grad_norm": 1.59375,
      "learning_rate": 5.036276733151392e-05,
      "loss": 0.033490809798240664,
      "memory(GiB)": 118.47,
      "step": 1055,
      "token_acc": 0.9894907033144705,
      "train_speed(iter/s)": 0.020909
    },
    {
      "epoch": 2.747566515249838,
      "grad_norm": 1.2421875,
      "learning_rate": 4.990930742116287e-05,
      "loss": 0.03164244592189789,
      "memory(GiB)": 118.47,
      "step": 1060,
      "token_acc": 0.9892915980230642,
      "train_speed(iter/s)": 0.020906
    },
    {
      "epoch": 2.7605451005840362,
      "grad_norm": 0.466796875,
      "learning_rate": 4.945585497036396e-05,
      "loss": 0.03462514877319336,
      "memory(GiB)": 118.47,
      "step": 1065,
      "token_acc": 0.9894050529747351,
      "train_speed(iter/s)": 0.020908
    },
    {
      "epoch": 2.773523685918235,
      "grad_norm": 0.640625,
      "learning_rate": 4.900244727601882e-05,
      "loss": 0.026110872626304626,
      "memory(GiB)": 118.47,
      "step": 1070,
      "token_acc": 0.9886086248982913,
      "train_speed(iter/s)": 0.020906
    },
    {
      "epoch": 2.7865022712524334,
      "grad_norm": 0.470703125,
      "learning_rate": 4.854912163134788e-05,
      "loss": 0.025288578867912293,
      "memory(GiB)": 118.47,
      "step": 1075,
      "token_acc": 0.9925496688741722,
      "train_speed(iter/s)": 0.020904
    },
    {
      "epoch": 2.799480856586632,
      "grad_norm": 0.63671875,
      "learning_rate": 4.809591532282282e-05,
      "loss": 0.017070406675338747,
      "memory(GiB)": 118.47,
      "step": 1080,
      "token_acc": 0.9924874791318865,
      "train_speed(iter/s)": 0.020904
    },
    {
      "epoch": 2.8124594419208306,
      "grad_norm": 2.03125,
      "learning_rate": 4.76428656270999e-05,
      "loss": 0.0368634819984436,
      "memory(GiB)": 118.47,
      "step": 1085,
      "token_acc": 0.9875724937862469,
      "train_speed(iter/s)": 0.020902
    },
    {
      "epoch": 2.8254380272550295,
      "grad_norm": 1.5546875,
      "learning_rate": 4.719000980795375e-05,
      "loss": 0.04762330949306488,
      "memory(GiB)": 118.47,
      "step": 1090,
      "token_acc": 0.9873843566021867,
      "train_speed(iter/s)": 0.020899
    },
    {
      "epoch": 2.838416612589228,
      "grad_norm": 1.28125,
      "learning_rate": 4.6737385113212516e-05,
      "loss": 0.03311816155910492,
      "memory(GiB)": 118.47,
      "step": 1095,
      "token_acc": 0.990139687756779,
      "train_speed(iter/s)": 0.020899
    },
    {
      "epoch": 2.851395197923426,
      "grad_norm": 1.078125,
      "learning_rate": 4.628502877169413e-05,
      "loss": 0.06551986932754517,
      "memory(GiB)": 118.47,
      "step": 1100,
      "token_acc": 0.9826589595375722,
      "train_speed(iter/s)": 0.020901
    },
    {
      "epoch": 2.864373783257625,
      "grad_norm": 1.203125,
      "learning_rate": 4.583297799014416e-05,
      "loss": 0.042209646105766295,
      "memory(GiB)": 118.47,
      "step": 1105,
      "token_acc": 0.9876543209876543,
      "train_speed(iter/s)": 0.020908
    },
    {
      "epoch": 2.8773523685918234,
      "grad_norm": 0.412109375,
      "learning_rate": 4.538126995017566e-05,
      "loss": 0.031407338380813596,
      "memory(GiB)": 118.47,
      "step": 1110,
      "token_acc": 0.988255033557047,
      "train_speed(iter/s)": 0.02091
    },
    {
      "epoch": 2.8903309539260222,
      "grad_norm": 1.46875,
      "learning_rate": 4.492994180521077e-05,
      "loss": 0.05650691390037536,
      "memory(GiB)": 118.47,
      "step": 1115,
      "token_acc": 0.9760726072607261,
      "train_speed(iter/s)": 0.020913
    },
    {
      "epoch": 2.9033095392602206,
      "grad_norm": 0.875,
      "learning_rate": 4.447903067742496e-05,
      "loss": 0.03302388489246368,
      "memory(GiB)": 118.47,
      "step": 1120,
      "token_acc": 0.988663967611336,
      "train_speed(iter/s)": 0.020919
    },
    {
      "epoch": 2.916288124594419,
      "grad_norm": 1.0546875,
      "learning_rate": 4.402857365469364e-05,
      "loss": 0.043651098012924196,
      "memory(GiB)": 118.47,
      "step": 1125,
      "token_acc": 0.986960065199674,
      "train_speed(iter/s)": 0.020924
    },
    {
      "epoch": 2.929266709928618,
      "grad_norm": 0.490234375,
      "learning_rate": 4.357860778754157e-05,
      "loss": 0.0316637396812439,
      "memory(GiB)": 118.47,
      "step": 1130,
      "token_acc": 0.9907016060862215,
      "train_speed(iter/s)": 0.020924
    },
    {
      "epoch": 2.942245295262816,
      "grad_norm": 0.79296875,
      "learning_rate": 4.3129170086095564e-05,
      "loss": 0.032447722554206845,
      "memory(GiB)": 118.47,
      "step": 1135,
      "token_acc": 0.9874266554903605,
      "train_speed(iter/s)": 0.020925
    },
    {
      "epoch": 2.955223880597015,
      "grad_norm": 1.296875,
      "learning_rate": 4.2680297517040285e-05,
      "loss": 0.026056140661239624,
      "memory(GiB)": 118.47,
      "step": 1140,
      "token_acc": 0.989344262295082,
      "train_speed(iter/s)": 0.020921
    },
    {
      "epoch": 2.9682024659312134,
      "grad_norm": 1.1015625,
      "learning_rate": 4.223202700057765e-05,
      "loss": 0.024412302672863005,
      "memory(GiB)": 118.47,
      "step": 1145,
      "token_acc": 0.9907795473595976,
      "train_speed(iter/s)": 0.020915
    },
    {
      "epoch": 2.981181051265412,
      "grad_norm": 1.546875,
      "learning_rate": 4.178439540739026e-05,
      "loss": 0.01829095035791397,
      "memory(GiB)": 118.47,
      "step": 1150,
      "token_acc": 0.991876523151909,
      "train_speed(iter/s)": 0.020917
    },
    {
      "epoch": 2.9941596365996106,
      "grad_norm": 0.3984375,
      "learning_rate": 4.133743955560857e-05,
      "loss": 0.024611912667751312,
      "memory(GiB)": 118.47,
      "step": 1155,
      "token_acc": 0.9909015715467329,
      "train_speed(iter/s)": 0.020918
    },
    {
      "epoch": 3.0051914341336796,
      "grad_norm": 0.4765625,
      "learning_rate": 4.0891196207782726e-05,
      "loss": 0.020613734424114228,
      "memory(GiB)": 118.49,
      "step": 1160,
      "token_acc": 0.9933206106870229,
      "train_speed(iter/s)": 0.020936
    },
    {
      "epoch": 3.018170019467878,
      "grad_norm": 1.0703125,
      "learning_rate": 4.0445702067858734e-05,
      "loss": 0.021938510239124298,
      "memory(GiB)": 118.49,
      "step": 1165,
      "token_acc": 0.9925681255161024,
      "train_speed(iter/s)": 0.020936
    },
    {
      "epoch": 3.0311486048020764,
      "grad_norm": 0.56640625,
      "learning_rate": 4.0000993778159444e-05,
      "loss": 0.010195696353912353,
      "memory(GiB)": 118.49,
      "step": 1170,
      "token_acc": 0.9966832504145937,
      "train_speed(iter/s)": 0.020933
    },
    {
      "epoch": 3.044127190136275,
      "grad_norm": 0.224609375,
      "learning_rate": 3.955710791637085e-05,
      "loss": 0.010604656487703323,
      "memory(GiB)": 118.49,
      "step": 1175,
      "token_acc": 0.9966216216216216,
      "train_speed(iter/s)": 0.020933
    },
    {
      "epoch": 3.0571057754704736,
      "grad_norm": 0.38671875,
      "learning_rate": 3.9114080992533445e-05,
      "loss": 0.021324822306632997,
      "memory(GiB)": 118.49,
      "step": 1180,
      "token_acc": 0.9967239967239967,
      "train_speed(iter/s)": 0.020935
    },
    {
      "epoch": 3.0700843608046724,
      "grad_norm": 0.09814453125,
      "learning_rate": 3.86719494460392e-05,
      "loss": 0.01656488925218582,
      "memory(GiB)": 118.49,
      "step": 1185,
      "token_acc": 0.9967186218211649,
      "train_speed(iter/s)": 0.020935
    },
    {
      "epoch": 3.0830629461388708,
      "grad_norm": 0.51171875,
      "learning_rate": 3.82307496426345e-05,
      "loss": 0.019095677137374877,
      "memory(GiB)": 118.49,
      "step": 1190,
      "token_acc": 0.9923534409515717,
      "train_speed(iter/s)": 0.020933
    },
    {
      "epoch": 3.0960415314730696,
      "grad_norm": 0.61328125,
      "learning_rate": 3.7790517871428895e-05,
      "loss": 0.009211428463459015,
      "memory(GiB)": 118.49,
      "step": 1195,
      "token_acc": 0.9958949096880131,
      "train_speed(iter/s)": 0.02093
    },
    {
      "epoch": 3.109020116807268,
      "grad_norm": 1.578125,
      "learning_rate": 3.7351290341910394e-05,
      "loss": 0.026492729783058167,
      "memory(GiB)": 118.49,
      "step": 1200,
      "token_acc": 0.9933665008291874,
      "train_speed(iter/s)": 0.020932
    },
    {
      "epoch": 3.1219987021414664,
      "grad_norm": 0.66015625,
      "learning_rate": 3.691310318096719e-05,
      "loss": 0.010792587697505952,
      "memory(GiB)": 118.49,
      "step": 1205,
      "token_acc": 0.9958333333333333,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.134977287475665,
      "grad_norm": 1.3515625,
      "learning_rate": 3.647599242991607e-05,
      "loss": 0.01512954980134964,
      "memory(GiB)": 118.49,
      "step": 1210,
      "token_acc": 0.9942763695829926,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.1479558728098636,
      "grad_norm": 0.921875,
      "learning_rate": 3.603999404153814e-05,
      "loss": 0.014608421921730041,
      "memory(GiB)": 118.49,
      "step": 1215,
      "token_acc": 0.9949748743718593,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.1609344581440624,
      "grad_norm": 0.6875,
      "learning_rate": 3.560514387712156e-05,
      "loss": 0.010922598093748093,
      "memory(GiB)": 118.49,
      "step": 1220,
      "token_acc": 0.996694214876033,
      "train_speed(iter/s)": 0.020931
    },
    {
      "epoch": 3.1739130434782608,
      "grad_norm": 0.828125,
      "learning_rate": 3.517147770351199e-05,
      "loss": 0.025766593217849732,
      "memory(GiB)": 118.49,
      "step": 1225,
      "token_acc": 0.992671009771987,
      "train_speed(iter/s)": 0.020933
    },
    {
      "epoch": 3.1868916288124596,
      "grad_norm": 0.486328125,
      "learning_rate": 3.473903119017071e-05,
      "loss": 0.01006721258163452,
      "memory(GiB)": 118.49,
      "step": 1230,
      "token_acc": 0.995928338762215,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.199870214146658,
      "grad_norm": 1.3203125,
      "learning_rate": 3.430783990624071e-05,
      "loss": 0.02418680191040039,
      "memory(GiB)": 118.49,
      "step": 1235,
      "token_acc": 0.9934264585045193,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.2128487994808568,
      "grad_norm": 0.33984375,
      "learning_rate": 3.3877939317621224e-05,
      "loss": 0.020583410561084748,
      "memory(GiB)": 118.49,
      "step": 1240,
      "token_acc": 0.9924874791318865,
      "train_speed(iter/s)": 0.020941
    },
    {
      "epoch": 3.225827384815055,
      "grad_norm": 1.359375,
      "learning_rate": 3.344936478405051e-05,
      "loss": 0.03402402400970459,
      "memory(GiB)": 118.49,
      "step": 1245,
      "token_acc": 0.9892915980230642,
      "train_speed(iter/s)": 0.02094
    },
    {
      "epoch": 3.2388059701492535,
      "grad_norm": 0.29296875,
      "learning_rate": 3.302215155619752e-05,
      "loss": 0.011006417870521545,
      "memory(GiB)": 118.49,
      "step": 1250,
      "token_acc": 0.9958711808422791,
      "train_speed(iter/s)": 0.020942
    },
    {
      "epoch": 3.2517845554834524,
      "grad_norm": 0.64453125,
      "learning_rate": 3.259633477276251e-05,
      "loss": 0.015421295166015625,
      "memory(GiB)": 118.49,
      "step": 1255,
      "token_acc": 0.9951060358890701,
      "train_speed(iter/s)": 0.020945
    },
    {
      "epoch": 3.2647631408176507,
      "grad_norm": 0.349609375,
      "learning_rate": 3.217194945758678e-05,
      "loss": 0.007819992303848267,
      "memory(GiB)": 118.49,
      "step": 1260,
      "token_acc": 0.9975267930750206,
      "train_speed(iter/s)": 0.020948
    },
    {
      "epoch": 3.2777417261518496,
      "grad_norm": 0.50390625,
      "learning_rate": 3.174903051677208e-05,
      "loss": 0.0145877406001091,
      "memory(GiB)": 118.49,
      "step": 1265,
      "token_acc": 0.9958779884583677,
      "train_speed(iter/s)": 0.020954
    },
    {
      "epoch": 3.290720311486048,
      "grad_norm": 0.84375,
      "learning_rate": 3.132761273580942e-05,
      "loss": 0.019092942774295806,
      "memory(GiB)": 118.49,
      "step": 1270,
      "token_acc": 0.992462311557789,
      "train_speed(iter/s)": 0.020953
    },
    {
      "epoch": 3.3036988968202468,
      "grad_norm": 0.90625,
      "learning_rate": 3.0907730776717894e-05,
      "loss": 0.019785448908805847,
      "memory(GiB)": 118.49,
      "step": 1275,
      "token_acc": 0.9926047658175843,
      "train_speed(iter/s)": 0.020952
    },
    {
      "epoch": 3.316677482154445,
      "grad_norm": 0.26953125,
      "learning_rate": 3.0489419175193866e-05,
      "loss": 0.009959718585014344,
      "memory(GiB)": 118.49,
      "step": 1280,
      "token_acc": 0.9958982772764561,
      "train_speed(iter/s)": 0.020951
    },
    {
      "epoch": 3.329656067488644,
      "grad_norm": 0.703125,
      "learning_rate": 3.0072712337770197e-05,
      "loss": 0.007476700097322464,
      "memory(GiB)": 118.49,
      "step": 1285,
      "token_acc": 0.9991568296795953,
      "train_speed(iter/s)": 0.020953
    },
    {
      "epoch": 3.3426346528228423,
      "grad_norm": 0.515625,
      "learning_rate": 2.96576445389864e-05,
      "loss": 0.004897921904921531,
      "memory(GiB)": 118.49,
      "step": 1290,
      "token_acc": 0.9975308641975309,
      "train_speed(iter/s)": 0.020954
    },
    {
      "epoch": 3.3556132381570407,
      "grad_norm": 0.65234375,
      "learning_rate": 2.924424991856947e-05,
      "loss": 0.016269049048423766,
      "memory(GiB)": 118.49,
      "step": 1295,
      "token_acc": 0.99581589958159,
      "train_speed(iter/s)": 0.020954
    },
    {
      "epoch": 3.3685918234912395,
      "grad_norm": 1.421875,
      "learning_rate": 2.8832562478625825e-05,
      "loss": 0.016820259392261505,
      "memory(GiB)": 118.49,
      "step": 1300,
      "token_acc": 0.9916597164303587,
      "train_speed(iter/s)": 0.020953
    },
    {
      "epoch": 3.381570408825438,
      "grad_norm": 0.458984375,
      "learning_rate": 2.84226160808447e-05,
      "loss": 0.016832391917705535,
      "memory(GiB)": 118.49,
      "step": 1305,
      "token_acc": 0.9942291838417148,
      "train_speed(iter/s)": 0.020954
    },
    {
      "epoch": 3.3945489941596367,
      "grad_norm": 0.49609375,
      "learning_rate": 2.8014444443712918e-05,
      "loss": 0.009478585422039032,
      "memory(GiB)": 118.49,
      "step": 1310,
      "token_acc": 0.9959083469721768,
      "train_speed(iter/s)": 0.020958
    },
    {
      "epoch": 3.407527579493835,
      "grad_norm": 0.68359375,
      "learning_rate": 2.7608081139741482e-05,
      "loss": 0.014081387221813202,
      "memory(GiB)": 118.49,
      "step": 1315,
      "token_acc": 0.9958711808422791,
      "train_speed(iter/s)": 0.020956
    },
    {
      "epoch": 3.420506164828034,
      "grad_norm": 0.79296875,
      "learning_rate": 2.7203559592704315e-05,
      "loss": 0.02275892049074173,
      "memory(GiB)": 118.49,
      "step": 1320,
      "token_acc": 0.9916597164303587,
      "train_speed(iter/s)": 0.020956
    },
    {
      "epoch": 3.4334847501622323,
      "grad_norm": 1.0546875,
      "learning_rate": 2.6800913074888983e-05,
      "loss": 0.011998300999403,
      "memory(GiB)": 118.49,
      "step": 1325,
      "token_acc": 0.9959415584415584,
      "train_speed(iter/s)": 0.020958
    },
    {
      "epoch": 3.446463335496431,
      "grad_norm": 0.6015625,
      "learning_rate": 2.6400174704360148e-05,
      "loss": 0.016256968677043914,
      "memory(GiB)": 118.49,
      "step": 1330,
      "token_acc": 0.9941569282136895,
      "train_speed(iter/s)": 0.020955
    },
    {
      "epoch": 3.4594419208306295,
      "grad_norm": 0.130859375,
      "learning_rate": 2.600137744223552e-05,
      "loss": 0.007994767278432846,
      "memory(GiB)": 118.49,
      "step": 1335,
      "token_acc": 0.9983361064891847,
      "train_speed(iter/s)": 0.020952
    },
    {
      "epoch": 3.472420506164828,
      "grad_norm": 0.8203125,
      "learning_rate": 2.5604554089974692e-05,
      "loss": 0.018077975511550902,
      "memory(GiB)": 118.49,
      "step": 1340,
      "token_acc": 0.9942434210526315,
      "train_speed(iter/s)": 0.020956
    },
    {
      "epoch": 3.4853990914990267,
      "grad_norm": 0.435546875,
      "learning_rate": 2.520973728668137e-05,
      "loss": 0.008464429527521133,
      "memory(GiB)": 118.49,
      "step": 1345,
      "token_acc": 0.9975470155355682,
      "train_speed(iter/s)": 0.020958
    },
    {
      "epoch": 3.498377676833225,
      "grad_norm": 0.392578125,
      "learning_rate": 2.4816959506418584e-05,
      "loss": 0.011597975343465804,
      "memory(GiB)": 118.49,
      "step": 1350,
      "token_acc": 0.9959016393442623,
      "train_speed(iter/s)": 0.02096
    },
    {
      "epoch": 3.511356262167424,
      "grad_norm": 0.32421875,
      "learning_rate": 2.442625305553779e-05,
      "loss": 0.013302412629127503,
      "memory(GiB)": 118.49,
      "step": 1355,
      "token_acc": 0.9957912457912458,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.5243348475016223,
      "grad_norm": 0.58984375,
      "learning_rate": 2.4037650070021624e-05,
      "loss": 0.010544488579034806,
      "memory(GiB)": 118.49,
      "step": 1360,
      "token_acc": 0.9975267930750206,
      "train_speed(iter/s)": 0.02096
    },
    {
      "epoch": 3.5373134328358207,
      "grad_norm": 0.7109375,
      "learning_rate": 2.3651182512840603e-05,
      "loss": 0.011504454165697097,
      "memory(GiB)": 118.49,
      "step": 1365,
      "token_acc": 0.9958915365653246,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.5502920181700195,
      "grad_norm": 0.7109375,
      "learning_rate": 2.3266882171324272e-05,
      "loss": 0.016246941685676575,
      "memory(GiB)": 118.49,
      "step": 1370,
      "token_acc": 0.9942622950819672,
      "train_speed(iter/s)": 0.020961
    },
    {
      "epoch": 3.5632706035042183,
      "grad_norm": 0.6640625,
      "learning_rate": 2.2884780654546584e-05,
      "loss": 0.005815648287534714,
      "memory(GiB)": 118.49,
      "step": 1375,
      "token_acc": 0.9983766233766234,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.5762491888384167,
      "grad_norm": 0.451171875,
      "learning_rate": 2.2504909390725975e-05,
      "loss": 0.009234168380498887,
      "memory(GiB)": 118.49,
      "step": 1380,
      "token_acc": 0.9975728155339806,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.589227774172615,
      "grad_norm": 0.265625,
      "learning_rate": 2.2127299624640508e-05,
      "loss": 0.004879490658640862,
      "memory(GiB)": 118.49,
      "step": 1385,
      "token_acc": 0.9983525535420099,
      "train_speed(iter/s)": 0.020965
    },
    {
      "epoch": 3.602206359506814,
      "grad_norm": 0.007049560546875,
      "learning_rate": 2.1751982415057798e-05,
      "loss": 0.007252056151628494,
      "memory(GiB)": 118.49,
      "step": 1390,
      "token_acc": 0.9967871485943776,
      "train_speed(iter/s)": 0.020963
    },
    {
      "epoch": 3.6151849448410123,
      "grad_norm": 0.08349609375,
      "learning_rate": 2.1378988632180523e-05,
      "loss": 0.006177261471748352,
      "memory(GiB)": 118.49,
      "step": 1395,
      "token_acc": 0.9975609756097561,
      "train_speed(iter/s)": 0.020959
    },
    {
      "epoch": 3.628163530175211,
      "grad_norm": 1.1796875,
      "learning_rate": 2.100834895510729e-05,
      "loss": 0.034820249676704405,
      "memory(GiB)": 118.49,
      "step": 1400,
      "token_acc": 0.992462311557789,
      "train_speed(iter/s)": 0.020959
    },
    {
      "epoch": 3.6411421155094095,
      "grad_norm": 0.83984375,
      "learning_rate": 2.064009386930915e-05,
      "loss": 0.022104519605636596,
      "memory(GiB)": 118.49,
      "step": 1405,
      "token_acc": 0.9942528735632183,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.654120700843608,
      "grad_norm": 0.89453125,
      "learning_rate": 2.0274253664122316e-05,
      "loss": 0.009468887746334077,
      "memory(GiB)": 118.49,
      "step": 1410,
      "token_acc": 0.9950041631973355,
      "train_speed(iter/s)": 0.020963
    },
    {
      "epoch": 3.6670992861778067,
      "grad_norm": 0.484375,
      "learning_rate": 1.991085843025667e-05,
      "loss": 0.01620858609676361,
      "memory(GiB)": 118.49,
      "step": 1415,
      "token_acc": 0.9950083194675541,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.680077871512005,
      "grad_norm": 0.671875,
      "learning_rate": 1.9549938057320893e-05,
      "loss": 0.013029226660728454,
      "memory(GiB)": 118.49,
      "step": 1420,
      "token_acc": 0.9934102141680395,
      "train_speed(iter/s)": 0.020967
    },
    {
      "epoch": 3.693056456846204,
      "grad_norm": 0.08447265625,
      "learning_rate": 1.9191522231363913e-05,
      "loss": 0.01667369157075882,
      "memory(GiB)": 118.49,
      "step": 1425,
      "token_acc": 0.9950657894736842,
      "train_speed(iter/s)": 0.020966
    },
    {
      "epoch": 3.7060350421804023,
      "grad_norm": 0.20703125,
      "learning_rate": 1.883564043243332e-05,
      "loss": 0.010355699062347411,
      "memory(GiB)": 118.49,
      "step": 1430,
      "token_acc": 0.9950940310711366,
      "train_speed(iter/s)": 0.020966
    },
    {
      "epoch": 3.719013627514601,
      "grad_norm": 1.8125,
      "learning_rate": 1.8482321932150467e-05,
      "loss": 0.011119000613689423,
      "memory(GiB)": 118.49,
      "step": 1435,
      "token_acc": 0.9934959349593496,
      "train_speed(iter/s)": 0.020968
    },
    {
      "epoch": 3.7319922128487995,
      "grad_norm": 0.08642578125,
      "learning_rate": 1.8131595791302974e-05,
      "loss": 0.007310156524181366,
      "memory(GiB)": 118.49,
      "step": 1440,
      "token_acc": 0.9958368026644463,
      "train_speed(iter/s)": 0.020965
    },
    {
      "epoch": 3.7449707981829983,
      "grad_norm": 0.2470703125,
      "learning_rate": 1.7783490857454356e-05,
      "loss": 0.005807420611381531,
      "memory(GiB)": 118.49,
      "step": 1445,
      "token_acc": 0.9983179142136249,
      "train_speed(iter/s)": 0.020965
    },
    {
      "epoch": 3.7579493835171967,
      "grad_norm": 0.890625,
      "learning_rate": 1.743803576257136e-05,
      "loss": 0.012262566387653351,
      "memory(GiB)": 118.49,
      "step": 1450,
      "token_acc": 0.9958949096880131,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.770927968851395,
      "grad_norm": 0.55859375,
      "learning_rate": 1.7095258920668866e-05,
      "loss": 0.004218662902712822,
      "memory(GiB)": 118.49,
      "step": 1455,
      "token_acc": 0.9991843393148451,
      "train_speed(iter/s)": 0.020966
    },
    {
      "epoch": 3.783906554185594,
      "grad_norm": 0.21875,
      "learning_rate": 1.6755188525472927e-05,
      "loss": 0.0074700690805912014,
      "memory(GiB)": 118.49,
      "step": 1460,
      "token_acc": 0.9991735537190083,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.7968851395197922,
      "grad_norm": 0.271484375,
      "learning_rate": 1.641785254810172e-05,
      "loss": 0.013314425945281982,
      "memory(GiB)": 118.49,
      "step": 1465,
      "token_acc": 0.9966527196652719,
      "train_speed(iter/s)": 0.020963
    },
    {
      "epoch": 3.809863724853991,
      "grad_norm": 0.54296875,
      "learning_rate": 1.608327873476488e-05,
      "loss": 0.008113903552293777,
      "memory(GiB)": 118.49,
      "step": 1470,
      "token_acc": 0.9983150800336984,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.8228423101881894,
      "grad_norm": 0.3828125,
      "learning_rate": 1.5751494604481427e-05,
      "loss": 0.008881012350320816,
      "memory(GiB)": 118.49,
      "step": 1475,
      "token_acc": 0.9950289975144988,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.835820895522388,
      "grad_norm": 0.76171875,
      "learning_rate": 1.542252744681627e-05,
      "loss": 0.00938112884759903,
      "memory(GiB)": 118.49,
      "step": 1480,
      "token_acc": 0.9967611336032388,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.8487994808565866,
      "grad_norm": 0.6015625,
      "learning_rate": 1.5096404319635532e-05,
      "loss": 0.01805244982242584,
      "memory(GiB)": 118.49,
      "step": 1485,
      "token_acc": 0.9935794542536116,
      "train_speed(iter/s)": 0.020963
    },
    {
      "epoch": 3.8617780661907855,
      "grad_norm": 0.58984375,
      "learning_rate": 1.4773152046881178e-05,
      "loss": 0.009089920669794083,
      "memory(GiB)": 118.49,
      "step": 1490,
      "token_acc": 0.997504159733777,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.874756651524984,
      "grad_norm": 0.1494140625,
      "learning_rate": 1.4452797216364527e-05,
      "loss": 0.011903107911348344,
      "memory(GiB)": 118.49,
      "step": 1495,
      "token_acc": 0.9974979149291076,
      "train_speed(iter/s)": 0.020962
    },
    {
      "epoch": 3.887735236859182,
      "grad_norm": 1.28125,
      "learning_rate": 1.4135366177579579e-05,
      "loss": 0.008489710092544556,
      "memory(GiB)": 118.49,
      "step": 1500,
      "token_acc": 0.9967266775777414,
      "train_speed(iter/s)": 0.020964
    },
    {
      "epoch": 3.900713822193381,
      "grad_norm": 0.07421875,
      "learning_rate": 1.3820885039535563e-05,
      "loss": 0.00586327463388443,
      "memory(GiB)": 118.49,
      "step": 1505,
      "token_acc": 0.9983818770226537,
      "train_speed(iter/s)": 0.020943
    },
    {
      "epoch": 3.9136924075275794,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.3509379668609545e-05,
      "loss": 0.0139107346534729,
      "memory(GiB)": 118.49,
      "step": 1510,
      "token_acc": 0.9917149958574979,
      "train_speed(iter/s)": 0.020943
    },
    {
      "epoch": 3.9266709928617782,
      "grad_norm": 0.6640625,
      "learning_rate": 1.3200875686418906e-05,
      "loss": 0.007857438921928406,
      "memory(GiB)": 118.49,
      "step": 1515,
      "token_acc": 0.9975409836065574,
      "train_speed(iter/s)": 0.020935
    },
    {
      "epoch": 3.9396495781959766,
      "grad_norm": 0.296875,
      "learning_rate": 1.289539846771387e-05,
      "loss": 0.014272244274616241,
      "memory(GiB)": 118.49,
      "step": 1520,
      "token_acc": 0.99597747385358,
      "train_speed(iter/s)": 0.020937
    },
    {
      "epoch": 3.952628163530175,
      "grad_norm": 0.10888671875,
      "learning_rate": 1.2592973138290459e-05,
      "loss": 0.011531826108694077,
      "memory(GiB)": 118.49,
      "step": 1525,
      "token_acc": 0.9974811083123426,
      "train_speed(iter/s)": 0.020935
    },
    {
      "epoch": 3.965606748864374,
      "grad_norm": 0.412109375,
      "learning_rate": 1.229362457292388e-05,
      "loss": 0.005492057651281357,
      "memory(GiB)": 118.49,
      "step": 1530,
      "token_acc": 0.998371335504886,
      "train_speed(iter/s)": 0.02093
    },
    {
      "epoch": 3.9785853341985726,
      "grad_norm": 0.81640625,
      "learning_rate": 1.1997377393322461e-05,
      "loss": 0.011842301487922669,
      "memory(GiB)": 118.49,
      "step": 1535,
      "token_acc": 0.997545008183306,
      "train_speed(iter/s)": 0.020928
    },
    {
      "epoch": 3.991563919532771,
      "grad_norm": 0.302734375,
      "learning_rate": 1.1704255966102628e-05,
      "loss": 0.009899959713220597,
      "memory(GiB)": 118.49,
      "step": 1540,
      "token_acc": 0.9958609271523179,
      "train_speed(iter/s)": 0.020926
    },
    {
      "epoch": 4.00259571706684,
      "grad_norm": 0.130859375,
      "learning_rate": 1.1414284400784642e-05,
      "loss": 0.005935117974877358,
      "memory(GiB)": 118.49,
      "step": 1545,
      "token_acc": 0.9990328820116054,
      "train_speed(iter/s)": 0.020936
    },
    {
      "epoch": 4.015574302401038,
      "grad_norm": 0.05517578125,
      "learning_rate": 1.1127486547809534e-05,
      "loss": 0.001661362312734127,
      "memory(GiB)": 118.49,
      "step": 1550,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02093
    },
    {
      "epoch": 4.028552887735237,
      "grad_norm": 0.09423828125,
      "learning_rate": 1.0843885996577502e-05,
      "loss": 0.0012222844175994395,
      "memory(GiB)": 118.49,
      "step": 1555,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02093
    },
    {
      "epoch": 4.041531473069435,
      "grad_norm": 0.1494140625,
      "learning_rate": 1.0563506073507523e-05,
      "loss": 0.0020720928907394407,
      "memory(GiB)": 118.49,
      "step": 1560,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020927
    },
    {
      "epoch": 4.054510058403634,
      "grad_norm": 0.404296875,
      "learning_rate": 1.0286369840118858e-05,
      "loss": 0.003256053477525711,
      "memory(GiB)": 118.49,
      "step": 1565,
      "token_acc": 0.9984,
      "train_speed(iter/s)": 0.020925
    },
    {
      "epoch": 4.067488643737833,
      "grad_norm": 0.373046875,
      "learning_rate": 1.0012500091134148e-05,
      "loss": 0.0011426467448472978,
      "memory(GiB)": 118.49,
      "step": 1570,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020924
    },
    {
      "epoch": 4.080467229072031,
      "grad_norm": 0.59375,
      "learning_rate": 9.741919352604506e-06,
      "loss": 0.0045836478471755985,
      "memory(GiB)": 118.49,
      "step": 1575,
      "token_acc": 0.9991809991809992,
      "train_speed(iter/s)": 0.020917
    },
    {
      "epoch": 4.09344581440623,
      "grad_norm": 0.46484375,
      "learning_rate": 9.47464988005678e-06,
      "loss": 0.0034479245543479918,
      "memory(GiB)": 118.49,
      "step": 1580,
      "token_acc": 0.9991749174917491,
      "train_speed(iter/s)": 0.020909
    },
    {
      "epoch": 4.106424399740428,
      "grad_norm": 0.875,
      "learning_rate": 9.210713656663023e-06,
      "loss": 0.007430323213338852,
      "memory(GiB)": 118.49,
      "step": 1585,
      "token_acc": 0.9983235540653814,
      "train_speed(iter/s)": 0.020896
    },
    {
      "epoch": 4.119402985074627,
      "grad_norm": 0.083984375,
      "learning_rate": 8.950132391432275e-06,
      "loss": 0.002633281424641609,
      "memory(GiB)": 118.49,
      "step": 1590,
      "token_acc": 0.9991922455573505,
      "train_speed(iter/s)": 0.020886
    },
    {
      "epoch": 4.132381570408826,
      "grad_norm": 0.73828125,
      "learning_rate": 8.692927517425092e-06,
      "loss": 0.006271419674158096,
      "memory(GiB)": 118.49,
      "step": 1595,
      "token_acc": 0.9975409836065574,
      "train_speed(iter/s)": 0.020881
    },
    {
      "epoch": 4.145360155743024,
      "grad_norm": 0.05322265625,
      "learning_rate": 8.439120189990535e-06,
      "loss": 0.0012338412925601006,
      "memory(GiB)": 118.49,
      "step": 1600,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02087
    },
    {
      "epoch": 4.158338741077222,
      "grad_norm": 0.0537109375,
      "learning_rate": 8.188731285026219e-06,
      "loss": 0.003544669598340988,
      "memory(GiB)": 118.49,
      "step": 1605,
      "token_acc": 0.9991889699918897,
      "train_speed(iter/s)": 0.020869
    },
    {
      "epoch": 4.171317326411421,
      "grad_norm": 0.1943359375,
      "learning_rate": 7.94178139726121e-06,
      "loss": 0.003528529778122902,
      "memory(GiB)": 118.49,
      "step": 1610,
      "token_acc": 0.9991961414790996,
      "train_speed(iter/s)": 0.020868
    },
    {
      "epoch": 4.18429591174562,
      "grad_norm": 0.07568359375,
      "learning_rate": 7.698290838562062e-06,
      "loss": 0.003579762205481529,
      "memory(GiB)": 118.49,
      "step": 1615,
      "token_acc": 0.998326359832636,
      "train_speed(iter/s)": 0.020864
    },
    {
      "epoch": 4.197274497079818,
      "grad_norm": 0.0517578125,
      "learning_rate": 7.4582796362622105e-06,
      "loss": 0.004392120614647865,
      "memory(GiB)": 118.49,
      "step": 1620,
      "token_acc": 0.9991742361684558,
      "train_speed(iter/s)": 0.02086
    },
    {
      "epoch": 4.210253082414017,
      "grad_norm": 0.40625,
      "learning_rate": 7.2217675315146395e-06,
      "loss": 0.0038825396448373796,
      "memory(GiB)": 118.49,
      "step": 1625,
      "token_acc": 0.9991532599491956,
      "train_speed(iter/s)": 0.02086
    },
    {
      "epoch": 4.223231667748215,
      "grad_norm": 0.1630859375,
      "learning_rate": 6.9887739776681945e-06,
      "loss": 0.0008973809890449047,
      "memory(GiB)": 118.49,
      "step": 1630,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020866
    },
    {
      "epoch": 4.236210253082414,
      "grad_norm": 0.1748046875,
      "learning_rate": 6.759318138667542e-06,
      "loss": 0.006377002596855164,
      "memory(GiB)": 118.49,
      "step": 1635,
      "token_acc": 0.9991735537190083,
      "train_speed(iter/s)": 0.02087
    },
    {
      "epoch": 4.249188838416613,
      "grad_norm": 0.224609375,
      "learning_rate": 6.533418887476833e-06,
      "loss": 0.004097684845328331,
      "memory(GiB)": 118.49,
      "step": 1640,
      "token_acc": 0.9991721854304636,
      "train_speed(iter/s)": 0.020869
    },
    {
      "epoch": 4.262167423750811,
      "grad_norm": 0.294921875,
      "learning_rate": 6.3110948045274896e-06,
      "loss": 0.0022493613883852958,
      "memory(GiB)": 118.49,
      "step": 1645,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020866
    },
    {
      "epoch": 4.2751460090850095,
      "grad_norm": 0.0172119140625,
      "learning_rate": 6.092364176189857e-06,
      "loss": 0.0012288546189665795,
      "memory(GiB)": 118.49,
      "step": 1650,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020865
    },
    {
      "epoch": 4.288124594419208,
      "grad_norm": 0.142578125,
      "learning_rate": 5.877244993269209e-06,
      "loss": 0.0014213725924491882,
      "memory(GiB)": 118.49,
      "step": 1655,
      "token_acc": 0.9991850040749797,
      "train_speed(iter/s)": 0.020866
    },
    {
      "epoch": 4.301103179753407,
      "grad_norm": 0.0224609375,
      "learning_rate": 5.665754949525914e-06,
      "loss": 0.000797341950237751,
      "memory(GiB)": 118.49,
      "step": 1660,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020865
    },
    {
      "epoch": 4.3140817650876055,
      "grad_norm": 0.28125,
      "learning_rate": 5.457911440220154e-06,
      "loss": 0.003249218687415123,
      "memory(GiB)": 118.49,
      "step": 1665,
      "token_acc": 0.9991789819376026,
      "train_speed(iter/s)": 0.020867
    },
    {
      "epoch": 4.327060350421804,
      "grad_norm": 0.30859375,
      "learning_rate": 5.253731560681141e-06,
      "loss": 0.001958874985575676,
      "memory(GiB)": 118.49,
      "step": 1670,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020868
    },
    {
      "epoch": 4.340038935756002,
      "grad_norm": 0.2490234375,
      "learning_rate": 5.0532321049010065e-06,
      "loss": 0.0055487435311079025,
      "memory(GiB)": 118.49,
      "step": 1675,
      "token_acc": 0.9991755976916735,
      "train_speed(iter/s)": 0.02087
    },
    {
      "epoch": 4.353017521090202,
      "grad_norm": 0.228515625,
      "learning_rate": 4.856429564153453e-06,
      "loss": 0.006631163507699966,
      "memory(GiB)": 118.49,
      "step": 1680,
      "token_acc": 0.9983566146261298,
      "train_speed(iter/s)": 0.020869
    },
    {
      "epoch": 4.3659961064244,
      "grad_norm": 0.291015625,
      "learning_rate": 4.6633401256373886e-06,
      "loss": 0.0010011550970375539,
      "memory(GiB)": 118.49,
      "step": 1685,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020871
    },
    {
      "epoch": 4.378974691758598,
      "grad_norm": 0.0400390625,
      "learning_rate": 4.473979671145445e-06,
      "loss": 0.0011706288903951644,
      "memory(GiB)": 118.49,
      "step": 1690,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020872
    },
    {
      "epoch": 4.391953277092797,
      "grad_norm": 0.04248046875,
      "learning_rate": 4.288363775757737e-06,
      "loss": 0.0005182858556509018,
      "memory(GiB)": 118.49,
      "step": 1695,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020876
    },
    {
      "epoch": 4.404931862426995,
      "grad_norm": 0.494140625,
      "learning_rate": 4.106507706560792e-06,
      "loss": 0.005208690091967583,
      "memory(GiB)": 118.49,
      "step": 1700,
      "token_acc": 0.9991701244813278,
      "train_speed(iter/s)": 0.020878
    },
    {
      "epoch": 4.417910447761194,
      "grad_norm": 0.1796875,
      "learning_rate": 3.9284264213917735e-06,
      "loss": 0.0008781940676271915,
      "memory(GiB)": 118.49,
      "step": 1705,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020877
    },
    {
      "epoch": 4.430889033095393,
      "grad_norm": 0.67578125,
      "learning_rate": 3.7541345676082475e-06,
      "loss": 0.010159996896982193,
      "memory(GiB)": 118.49,
      "step": 1710,
      "token_acc": 0.9975247524752475,
      "train_speed(iter/s)": 0.020878
    },
    {
      "epoch": 4.443867618429591,
      "grad_norm": 0.0169677734375,
      "learning_rate": 3.5836464808833625e-06,
      "loss": 0.000509789539501071,
      "memory(GiB)": 118.49,
      "step": 1715,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020882
    },
    {
      "epoch": 4.4568462037637895,
      "grad_norm": 0.51953125,
      "learning_rate": 3.4169761840267824e-06,
      "loss": 0.0018530094996094704,
      "memory(GiB)": 118.49,
      "step": 1720,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020878
    },
    {
      "epoch": 4.469824789097988,
      "grad_norm": 0.36328125,
      "learning_rate": 3.2541373858312628e-06,
      "loss": 0.0017950931563973428,
      "memory(GiB)": 118.49,
      "step": 1725,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020874
    },
    {
      "epoch": 4.482803374432187,
      "grad_norm": 0.083984375,
      "learning_rate": 3.0951434799450784e-06,
      "loss": 0.001635444164276123,
      "memory(GiB)": 118.49,
      "step": 1730,
      "token_acc": 0.9991803278688525,
      "train_speed(iter/s)": 0.020874
    },
    {
      "epoch": 4.4957819597663855,
      "grad_norm": 0.28515625,
      "learning_rate": 2.9400075437704287e-06,
      "loss": 0.005332779884338379,
      "memory(GiB)": 118.49,
      "step": 1735,
      "token_acc": 0.998330550918197,
      "train_speed(iter/s)": 0.02087
    },
    {
      "epoch": 4.508760545100584,
      "grad_norm": 0.126953125,
      "learning_rate": 2.7887423373877673e-06,
      "loss": 0.008843471109867097,
      "memory(GiB)": 118.49,
      "step": 1740,
      "token_acc": 0.9975227085053675,
      "train_speed(iter/s)": 0.020865
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 0.10693359375,
      "learning_rate": 2.6413603025062792e-06,
      "loss": 0.0013348528183996678,
      "memory(GiB)": 118.49,
      "step": 1745,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02086
    },
    {
      "epoch": 4.5347177157689815,
      "grad_norm": 0.027587890625,
      "learning_rate": 2.4978735614405668e-06,
      "loss": 0.0018230179324746133,
      "memory(GiB)": 118.49,
      "step": 1750,
      "token_acc": 0.9991673605328892,
      "train_speed(iter/s)": 0.02086
    },
    {
      "epoch": 4.54769630110318,
      "grad_norm": 0.28515625,
      "learning_rate": 2.3582939161135353e-06,
      "loss": 0.0033437285572290422,
      "memory(GiB)": 118.49,
      "step": 1755,
      "token_acc": 0.9991610738255033,
      "train_speed(iter/s)": 0.020857
    },
    {
      "epoch": 4.560674886437378,
      "grad_norm": 0.1796875,
      "learning_rate": 2.222632847085715e-06,
      "loss": 0.001527964323759079,
      "memory(GiB)": 118.49,
      "step": 1760,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020854
    },
    {
      "epoch": 4.573653471771577,
      "grad_norm": 0.005401611328125,
      "learning_rate": 2.090901512610949e-06,
      "loss": 0.006669136136770249,
      "memory(GiB)": 118.49,
      "step": 1765,
      "token_acc": 0.9991783073130649,
      "train_speed(iter/s)": 0.020851
    },
    {
      "epoch": 4.586632057105776,
      "grad_norm": 0.00946044921875,
      "learning_rate": 1.9631107477185995e-06,
      "loss": 0.004949812591075897,
      "memory(GiB)": 118.49,
      "step": 1770,
      "token_acc": 0.9983471074380166,
      "train_speed(iter/s)": 0.020843
    },
    {
      "epoch": 4.599610642439974,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.8392710633224064e-06,
      "loss": 0.0032926712185144423,
      "memory(GiB)": 118.49,
      "step": 1775,
      "token_acc": 0.9983416252072969,
      "train_speed(iter/s)": 0.020848
    },
    {
      "epoch": 4.612589227774173,
      "grad_norm": 0.0262451171875,
      "learning_rate": 1.7193926453559272e-06,
      "loss": 0.0015199462883174419,
      "memory(GiB)": 118.49,
      "step": 1780,
      "token_acc": 0.9991830065359477,
      "train_speed(iter/s)": 0.020849
    },
    {
      "epoch": 4.625567813108371,
      "grad_norm": 0.42578125,
      "learning_rate": 1.6034853539347029e-06,
      "loss": 0.005044208467006683,
      "memory(GiB)": 118.49,
      "step": 1785,
      "token_acc": 0.9983646770237122,
      "train_speed(iter/s)": 0.020846
    },
    {
      "epoch": 4.638546398442569,
      "grad_norm": 0.045166015625,
      "learning_rate": 1.491558722545311e-06,
      "loss": 0.0026885632425546644,
      "memory(GiB)": 118.49,
      "step": 1790,
      "token_acc": 0.9983753046303818,
      "train_speed(iter/s)": 0.020847
    },
    {
      "epoch": 4.651524983776769,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.3836219572611886e-06,
      "loss": 0.001484157331287861,
      "memory(GiB)": 118.49,
      "step": 1795,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020847
    },
    {
      "epoch": 4.664503569110967,
      "grad_norm": 0.298828125,
      "learning_rate": 1.2796839359854374e-06,
      "loss": 0.0019209956750273704,
      "memory(GiB)": 118.49,
      "step": 1800,
      "token_acc": 0.9991708126036484,
      "train_speed(iter/s)": 0.020842
    },
    {
      "epoch": 4.6774821544451655,
      "grad_norm": 0.07861328125,
      "learning_rate": 1.1797532077206187e-06,
      "loss": 0.003924115002155304,
      "memory(GiB)": 118.49,
      "step": 1805,
      "token_acc": 0.9983498349834984,
      "train_speed(iter/s)": 0.020835
    },
    {
      "epoch": 4.690460739779364,
      "grad_norm": 0.240234375,
      "learning_rate": 1.0838379918655495e-06,
      "loss": 0.003906024992465973,
      "memory(GiB)": 118.49,
      "step": 1810,
      "token_acc": 0.9991721854304636,
      "train_speed(iter/s)": 0.02083
    },
    {
      "epoch": 4.703439325113562,
      "grad_norm": 0.0361328125,
      "learning_rate": 9.919461775392935e-07,
      "loss": 0.002249754220247269,
      "memory(GiB)": 118.49,
      "step": 1815,
      "token_acc": 0.9983593109105825,
      "train_speed(iter/s)": 0.020827
    },
    {
      "epoch": 4.7164179104477615,
      "grad_norm": 0.02880859375,
      "learning_rate": 9.040853229322577e-07,
      "loss": 0.0005314781330525876,
      "memory(GiB)": 118.49,
      "step": 1820,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020824
    },
    {
      "epoch": 4.72939649578196,
      "grad_norm": 0.06494140625,
      "learning_rate": 8.202626546845171e-07,
      "loss": 0.0008674643933773041,
      "memory(GiB)": 118.49,
      "step": 1825,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02082
    },
    {
      "epoch": 4.742375081116158,
      "grad_norm": 0.04150390625,
      "learning_rate": 7.404850672914243e-07,
      "loss": 0.0025060476735234262,
      "memory(GiB)": 118.49,
      "step": 1830,
      "token_acc": 0.9991341991341991,
      "train_speed(iter/s)": 0.020813
    },
    {
      "epoch": 4.755353666450357,
      "grad_norm": 0.007293701171875,
      "learning_rate": 6.647591225365235e-07,
      "loss": 0.0018194744363427163,
      "memory(GiB)": 118.49,
      "step": 1835,
      "token_acc": 0.9991902834008097,
      "train_speed(iter/s)": 0.020817
    },
    {
      "epoch": 4.768332251784555,
      "grad_norm": 0.037109375,
      "learning_rate": 5.930910489518493e-07,
      "loss": 0.0006959903985261917,
      "memory(GiB)": 118.49,
      "step": 1840,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020812
    },
    {
      "epoch": 4.781310837118754,
      "grad_norm": 0.5078125,
      "learning_rate": 5.254867413056197e-07,
      "loss": 0.0052085991948843,
      "memory(GiB)": 118.49,
      "step": 1845,
      "token_acc": 0.9983606557377049,
      "train_speed(iter/s)": 0.02081
    },
    {
      "epoch": 4.794289422452953,
      "grad_norm": 0.5625,
      "learning_rate": 4.619517601173573e-07,
      "loss": 0.005099215358495712,
      "memory(GiB)": 118.49,
      "step": 1850,
      "token_acc": 0.9991666666666666,
      "train_speed(iter/s)": 0.020811
    },
    {
      "epoch": 4.807268007787151,
      "grad_norm": 0.04150390625,
      "learning_rate": 4.0249133120059404e-07,
      "loss": 0.007123344391584396,
      "memory(GiB)": 118.49,
      "step": 1855,
      "token_acc": 0.9983593109105825,
      "train_speed(iter/s)": 0.020807
    },
    {
      "epoch": 4.820246593121349,
      "grad_norm": 0.0634765625,
      "learning_rate": 3.47110345232976e-07,
      "loss": 0.0038691792637109757,
      "memory(GiB)": 118.49,
      "step": 1860,
      "token_acc": 0.9983539094650206,
      "train_speed(iter/s)": 0.020812
    },
    {
      "epoch": 4.833225178455549,
      "grad_norm": 0.345703125,
      "learning_rate": 2.958133573540467e-07,
      "loss": 0.001323844026774168,
      "memory(GiB)": 118.49,
      "step": 1865,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020812
    },
    {
      "epoch": 4.846203763789747,
      "grad_norm": 0.119140625,
      "learning_rate": 2.486045867905573e-07,
      "loss": 0.00251984428614378,
      "memory(GiB)": 118.49,
      "step": 1870,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020811
    },
    {
      "epoch": 4.859182349123945,
      "grad_norm": 0.09375,
      "learning_rate": 2.0548791650944477e-07,
      "loss": 0.0011555158533155918,
      "memory(GiB)": 118.49,
      "step": 1875,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02081
    },
    {
      "epoch": 4.872160934458144,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.6646689289844254e-07,
      "loss": 0.002080231159925461,
      "memory(GiB)": 118.49,
      "step": 1880,
      "token_acc": 0.9991680532445923,
      "train_speed(iter/s)": 0.020807
    },
    {
      "epoch": 4.885139519792343,
      "grad_norm": 0.50390625,
      "learning_rate": 1.315447254744029e-07,
      "loss": 0.0027224332094192504,
      "memory(GiB)": 118.49,
      "step": 1885,
      "token_acc": 0.998351195383347,
      "train_speed(iter/s)": 0.020808
    },
    {
      "epoch": 4.898118105126541,
      "grad_norm": 0.326171875,
      "learning_rate": 1.0072428661929701e-07,
      "loss": 0.0020501598715782165,
      "memory(GiB)": 118.49,
      "step": 1890,
      "token_acc": 0.9991539763113367,
      "train_speed(iter/s)": 0.020809
    },
    {
      "epoch": 4.91109669046074,
      "grad_norm": 0.03515625,
      "learning_rate": 7.400811134395946e-08,
      "loss": 0.0011606317013502121,
      "memory(GiB)": 118.49,
      "step": 1895,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020811
    },
    {
      "epoch": 4.924075275794938,
      "grad_norm": 0.08935546875,
      "learning_rate": 5.139839707958283e-08,
      "loss": 0.0006340864114463329,
      "memory(GiB)": 118.49,
      "step": 1900,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.02081
    },
    {
      "epoch": 4.937053861129137,
      "grad_norm": 0.1357421875,
      "learning_rate": 3.289700349698999e-08,
      "loss": 0.00033661627676337955,
      "memory(GiB)": 118.49,
      "step": 1905,
      "token_acc": 1.0,
      "train_speed(iter/s)": 0.020812
    },
    {
      "epoch": 4.950032446463336,
      "grad_norm": 0.40625,
      "learning_rate": 1.8505452353656526e-08,
      "loss": 0.006044203415513039,
      "memory(GiB)": 118.49,
      "step": 1910,
      "token_acc": 0.9984051036682615,
      "train_speed(iter/s)": 0.020814
    },
    {
      "epoch": 4.963011031797534,
      "grad_norm": 0.0079345703125,
      "learning_rate": 8.22492736854974e-09,
      "loss": 0.00470067672431469,
      "memory(GiB)": 118.49,
      "step": 1915,
      "token_acc": 0.9983525535420099,
      "train_speed(iter/s)": 0.020815
    },
    {
      "epoch": 4.975989617131733,
      "grad_norm": 0.76171875,
      "learning_rate": 2.056274124773161e-09,
      "loss": 0.007594909518957138,
      "memory(GiB)": 118.49,
      "step": 1920,
      "token_acc": 0.9975186104218362,
      "train_speed(iter/s)": 0.020814
    },
    {
      "epoch": 4.988968202465931,
      "grad_norm": 0.0086669921875,
      "learning_rate": 0.0,
      "loss": 0.007923491299152374,
      "memory(GiB)": 118.49,
      "step": 1925,
      "token_acc": 0.9975062344139651,
      "train_speed(iter/s)": 0.020815
    }
  ],
  "logging_steps": 5,
  "max_steps": 1925,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.70560890425983e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}