{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5555555555555554,
  "eval_steps": 500,
  "global_step": 2001,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01277139208173691,
      "grad_norm": 1.7130959033966064,
      "learning_rate": 2.25e-06,
      "loss": 1.3267,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 10
    },
    {
      "epoch": 0.02554278416347382,
      "grad_norm": 2.028806209564209,
      "learning_rate": 4.75e-06,
      "loss": 0.8937,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 20
    },
    {
      "epoch": 0.038314176245210725,
      "grad_norm": 2.7790589332580566,
      "learning_rate": 7.25e-06,
      "loss": 0.7495,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 30
    },
    {
      "epoch": 0.05108556832694764,
      "grad_norm": 1.1010891199111938,
      "learning_rate": 9.750000000000002e-06,
      "loss": 0.688,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 40
    },
    {
      "epoch": 0.06385696040868455,
      "grad_norm": 0.694149911403656,
      "learning_rate": 1.225e-05,
      "loss": 0.6735,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 50
    },
    {
      "epoch": 0.07662835249042145,
      "grad_norm": 2.410159111022949,
      "learning_rate": 1.475e-05,
      "loss": 0.6393,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 60
    },
    {
      "epoch": 0.08939974457215837,
      "grad_norm": 0.560838520526886,
      "learning_rate": 1.725e-05,
      "loss": 0.6485,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 70
    },
    {
      "epoch": 0.10217113665389528,
      "grad_norm": 0.5268104076385498,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 0.6122,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 80
    },
    {
      "epoch": 0.11494252873563218,
      "grad_norm": 0.6267268657684326,
      "learning_rate": 2.2250000000000002e-05,
      "loss": 0.6448,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 90
    },
    {
      "epoch": 0.1277139208173691,
      "grad_norm": 0.5013255476951599,
      "learning_rate": 2.4750000000000002e-05,
      "loss": 0.6189,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 100
    },
    {
      "epoch": 0.140485312899106,
      "grad_norm": 0.4797821342945099,
      "learning_rate": 2.725e-05,
      "loss": 0.6158,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 110
    },
    {
      "epoch": 0.1532567049808429,
      "grad_norm": 0.5593216419219971,
      "learning_rate": 2.975e-05,
      "loss": 0.6299,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 120
    },
    {
      "epoch": 0.16602809706257982,
      "grad_norm": 0.6203834414482117,
      "learning_rate": 3.2250000000000005e-05,
      "loss": 0.6147,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 130
    },
    {
      "epoch": 0.17879948914431673,
      "grad_norm": 0.5475884675979614,
      "learning_rate": 3.475e-05,
      "loss": 0.6388,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 140
    },
    {
      "epoch": 0.19157088122605365,
      "grad_norm": 0.5112230777740479,
      "learning_rate": 3.7250000000000004e-05,
      "loss": 0.6078,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 150
    },
    {
      "epoch": 0.20434227330779056,
      "grad_norm": 0.5234256982803345,
      "learning_rate": 3.9750000000000004e-05,
      "loss": 0.6171,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 160
    },
    {
      "epoch": 0.21711366538952745,
      "grad_norm": 0.4764348566532135,
      "learning_rate": 4.2250000000000004e-05,
      "loss": 0.6129,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 170
    },
    {
      "epoch": 0.22988505747126436,
      "grad_norm": 0.4990110993385315,
      "learning_rate": 4.4750000000000004e-05,
      "loss": 0.6082,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 180
    },
    {
      "epoch": 0.24265644955300128,
      "grad_norm": 0.4930218756198883,
      "learning_rate": 4.7249999999999997e-05,
      "loss": 0.5917,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 190
    },
    {
      "epoch": 0.2554278416347382,
      "grad_norm": 0.5820630788803101,
      "learning_rate": 4.975e-05,
      "loss": 0.6155,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 200
    },
    {
      "epoch": 0.2681992337164751,
      "grad_norm": 0.7310104966163635,
      "learning_rate": 4.999784346519516e-05,
      "loss": 0.6128,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 210
    },
    {
      "epoch": 0.280970625798212,
      "grad_norm": 0.5114974975585938,
      "learning_rate": 4.999038945934662e-05,
      "loss": 0.6011,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 220
    },
    {
      "epoch": 0.2937420178799489,
      "grad_norm": 0.548307478427887,
      "learning_rate": 4.997761362606955e-05,
      "loss": 0.642,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 230
    },
    {
      "epoch": 0.3065134099616858,
      "grad_norm": 0.5858580470085144,
      "learning_rate": 4.995951985268617e-05,
      "loss": 0.643,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 240
    },
    {
      "epoch": 0.31928480204342274,
      "grad_norm": 0.5234502553939819,
      "learning_rate": 4.9936113644616454e-05,
      "loss": 0.646,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 250
    },
    {
      "epoch": 0.33205619412515963,
      "grad_norm": 0.7147364616394043,
      "learning_rate": 4.9907402123702983e-05,
      "loss": 0.6357,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 260
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.5782094597816467,
      "learning_rate": 4.987339402604392e-05,
      "loss": 0.6281,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 270
    },
    {
      "epoch": 0.35759897828863346,
      "grad_norm": 0.5938964486122131,
      "learning_rate": 4.9834099699334926e-05,
      "loss": 0.6516,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 280
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.6496431231498718,
      "learning_rate": 4.97895310997206e-05,
      "loss": 0.6183,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 290
    },
    {
      "epoch": 0.3831417624521073,
      "grad_norm": 0.7002574801445007,
      "learning_rate": 4.973970178815661e-05,
      "loss": 0.6441,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 300
    },
    {
      "epoch": 0.3959131545338442,
      "grad_norm": 0.6120892763137817,
      "learning_rate": 4.968462692628344e-05,
      "loss": 0.6456,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 310
    },
    {
      "epoch": 0.4086845466155811,
      "grad_norm": 0.6768554449081421,
      "learning_rate": 4.962432327181316e-05,
      "loss": 0.6373,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 320
    },
    {
      "epoch": 0.421455938697318,
      "grad_norm": 0.6195896863937378,
      "learning_rate": 4.955880917343052e-05,
      "loss": 0.6073,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 330
    },
    {
      "epoch": 0.4342273307790549,
      "grad_norm": 0.7980285882949829,
      "learning_rate": 4.9488104565209984e-05,
      "loss": 0.6247,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 340
    },
    {
      "epoch": 0.44699872286079184,
      "grad_norm": 0.6366240978240967,
      "learning_rate": 4.9412230960550334e-05,
      "loss": 0.6109,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 350
    },
    {
      "epoch": 0.45977011494252873,
      "grad_norm": 0.675498366355896,
      "learning_rate": 4.9331211445628766e-05,
      "loss": 0.5995,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 360
    },
    {
      "epoch": 0.4725415070242657,
      "grad_norm": 0.6284672617912292,
      "learning_rate": 4.9245070672376426e-05,
      "loss": 0.6389,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 370
    },
    {
      "epoch": 0.48531289910600256,
      "grad_norm": 0.6955174207687378,
      "learning_rate": 4.9153834850977556e-05,
      "loss": 0.6423,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 380
    },
    {
      "epoch": 0.49808429118773945,
      "grad_norm": 0.7838078141212463,
      "learning_rate": 4.9057531741894414e-05,
      "loss": 0.6231,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 390
    },
    {
      "epoch": 0.5108556832694764,
      "grad_norm": 0.7328115105628967,
      "learning_rate": 4.895619064742063e-05,
      "loss": 0.6109,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 400
    },
    {
      "epoch": 0.5236270753512133,
      "grad_norm": 0.7829299569129944,
      "learning_rate": 4.884984240276529e-05,
      "loss": 0.6317,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 410
    },
    {
      "epoch": 0.5363984674329502,
      "grad_norm": 0.7022182941436768,
      "learning_rate": 4.8738519366670696e-05,
      "loss": 0.6423,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 420
    },
    {
      "epoch": 0.5491698595146871,
      "grad_norm": 0.7584450840950012,
      "learning_rate": 4.862225541156653e-05,
      "loss": 0.6231,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 430
    },
    {
      "epoch": 0.561941251596424,
      "grad_norm": 0.8013651967048645,
      "learning_rate": 4.8501085913263424e-05,
      "loss": 0.6236,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 440
    },
    {
      "epoch": 0.5747126436781609,
      "grad_norm": 0.7047580480575562,
      "learning_rate": 4.83750477401891e-05,
      "loss": 0.613,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 450
    },
    {
      "epoch": 0.5874840357598978,
      "grad_norm": 0.6925296783447266,
      "learning_rate": 4.824417924217033e-05,
      "loss": 0.6024,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 460
    },
    {
      "epoch": 0.6002554278416348,
      "grad_norm": 0.7792147397994995,
      "learning_rate": 4.8108520238764266e-05,
      "loss": 0.6369,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 470
    },
    {
      "epoch": 0.6130268199233716,
      "grad_norm": 0.7831078171730042,
      "learning_rate": 4.7968112007142404e-05,
      "loss": 0.6276,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 480
    },
    {
      "epoch": 0.6257982120051085,
      "grad_norm": 0.7804124355316162,
      "learning_rate": 4.7822997269531154e-05,
      "loss": 0.6191,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 490
    },
    {
      "epoch": 0.6385696040868455,
      "grad_norm": 0.9669263958930969,
      "learning_rate": 4.767322018021263e-05,
      "loss": 0.605,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 500
    },
    {
      "epoch": 0.6513409961685823,
      "grad_norm": 0.7694301605224609,
      "learning_rate": 4.751882631208984e-05,
      "loss": 0.6559,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 510
    },
    {
      "epoch": 0.6641123882503193,
      "grad_norm": 0.7080773711204529,
      "learning_rate": 4.7359862642820076e-05,
      "loss": 0.6394,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 520
    },
    {
      "epoch": 0.6768837803320562,
      "grad_norm": 0.822318971157074,
      "learning_rate": 4.7196377540520995e-05,
      "loss": 0.6268,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 530
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.7512189745903015,
      "learning_rate": 4.702842074905359e-05,
      "loss": 0.6126,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 540
    },
    {
      "epoch": 0.70242656449553,
      "grad_norm": 0.7961385250091553,
      "learning_rate": 4.685604337288648e-05,
      "loss": 0.6205,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 550
    },
    {
      "epoch": 0.7151979565772669,
      "grad_norm": 0.7834734916687012,
      "learning_rate": 4.6679297861546366e-05,
      "loss": 0.6316,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 560
    },
    {
      "epoch": 0.7279693486590039,
      "grad_norm": 0.811787486076355,
      "learning_rate": 4.6498237993659036e-05,
      "loss": 0.614,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 570
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.8118691444396973,
      "learning_rate": 4.631291886058615e-05,
      "loss": 0.635,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 580
    },
    {
      "epoch": 0.7535121328224776,
      "grad_norm": 0.8244702219963074,
      "learning_rate": 4.612339684966247e-05,
      "loss": 0.6186,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 590
    },
    {
      "epoch": 0.7662835249042146,
      "grad_norm": 0.9031332731246948,
      "learning_rate": 4.5929729627038787e-05,
      "loss": 0.6371,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 600
    },
    {
      "epoch": 0.7790549169859514,
      "grad_norm": 0.8878704905509949,
      "learning_rate": 4.573197612013578e-05,
      "loss": 0.6421,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 610
    },
    {
      "epoch": 0.7918263090676884,
      "grad_norm": 0.8676999807357788,
      "learning_rate": 4.553019649971404e-05,
      "loss": 0.6163,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 620
    },
    {
      "epoch": 0.8045977011494253,
      "grad_norm": 0.8879650235176086,
      "learning_rate": 4.532445216156589e-05,
      "loss": 0.6116,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 630
    },
    {
      "epoch": 0.8173690932311622,
      "grad_norm": 0.8578729629516602,
      "learning_rate": 4.511480570783431e-05,
      "loss": 0.6461,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 640
    },
    {
      "epoch": 0.8301404853128991,
      "grad_norm": 0.8495433330535889,
      "learning_rate": 4.490132092796494e-05,
      "loss": 0.6164,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 650
    },
    {
      "epoch": 0.842911877394636,
      "grad_norm": 0.8818103075027466,
      "learning_rate": 4.46840627792968e-05,
      "loss": 0.645,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 660
    },
    {
      "epoch": 0.855683269476373,
      "grad_norm": 0.7751449942588806,
      "learning_rate": 4.446309736729751e-05,
      "loss": 0.6223,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 670
    },
    {
      "epoch": 0.8684546615581098,
      "grad_norm": 0.8533280491828918,
      "learning_rate": 4.4238491925449396e-05,
      "loss": 0.6108,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 680
    },
    {
      "epoch": 0.8812260536398467,
      "grad_norm": 0.7644783854484558,
      "learning_rate": 4.4010314794792196e-05,
      "loss": 0.5845,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 690
    },
    {
      "epoch": 0.8939974457215837,
      "grad_norm": 0.8928096890449524,
      "learning_rate": 4.377863540312882e-05,
      "loss": 0.6389,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 700
    },
    {
      "epoch": 0.9067688378033205,
      "grad_norm": 0.9339970350265503,
      "learning_rate": 4.35435242439005e-05,
      "loss": 0.5905,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 710
    },
    {
      "epoch": 0.9195402298850575,
      "grad_norm": 0.8004916310310364,
      "learning_rate": 4.3305052854737595e-05,
      "loss": 0.654,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 720
    },
    {
      "epoch": 0.9323116219667944,
      "grad_norm": 0.9015440940856934,
      "learning_rate": 4.306329379569273e-05,
      "loss": 0.634,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 730
    },
    {
      "epoch": 0.9450830140485313,
      "grad_norm": 0.8131055235862732,
      "learning_rate": 4.281832062716284e-05,
      "loss": 0.6431,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 740
    },
    {
      "epoch": 0.9578544061302682,
      "grad_norm": 0.8440934419631958,
      "learning_rate": 4.257020788750692e-05,
      "loss": 0.6495,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 750
    },
    {
      "epoch": 0.9706257982120051,
      "grad_norm": 0.8826728463172913,
      "learning_rate": 4.2319031070365945e-05,
      "loss": 0.6257,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 760
    },
    {
      "epoch": 0.9833971902937421,
      "grad_norm": 0.9320764541625977,
      "learning_rate": 4.206486660169249e-05,
      "loss": 0.5903,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 770
    },
    {
      "epoch": 0.9961685823754789,
      "grad_norm": 0.8958304524421692,
      "learning_rate": 4.180779181649638e-05,
      "loss": 0.6368,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 780
    },
    {
      "epoch": 1.0089399744572158,
      "grad_norm": 0.9672563076019287,
      "learning_rate": 4.154788493531388e-05,
      "loss": 0.5189,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 790
    },
    {
      "epoch": 1.0217113665389528,
      "grad_norm": 1.3082671165466309,
      "learning_rate": 4.1285225040407406e-05,
      "loss": 0.471,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 800
    },
    {
      "epoch": 1.0344827586206897,
      "grad_norm": 1.069593071937561,
      "learning_rate": 4.101989205170308e-05,
      "loss": 0.4711,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 810
    },
    {
      "epoch": 1.0472541507024267,
      "grad_norm": 1.0737265348434448,
      "learning_rate": 4.075196670247331e-05,
      "loss": 0.4945,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 820
    },
    {
      "epoch": 1.0600255427841634,
      "grad_norm": 0.9395204782485962,
      "learning_rate": 4.0481530514772006e-05,
      "loss": 0.4973,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 830
    },
    {
      "epoch": 1.0727969348659003,
      "grad_norm": 0.9196081757545471,
      "learning_rate": 4.0208665774629714e-05,
      "loss": 0.4798,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 840
    },
    {
      "epoch": 1.0855683269476373,
      "grad_norm": 1.0082074403762817,
      "learning_rate": 3.9933455507016334e-05,
      "loss": 0.5056,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 850
    },
    {
      "epoch": 1.0983397190293742,
      "grad_norm": 0.8791260719299316,
      "learning_rate": 3.9655983450578995e-05,
      "loss": 0.4971,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 860
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 1.105723261833191,
      "learning_rate": 3.937633403216271e-05,
      "loss": 0.5073,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 870
    },
    {
      "epoch": 1.123882503192848,
      "grad_norm": 0.9755126237869263,
      "learning_rate": 3.9094592341121786e-05,
      "loss": 0.5024,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 880
    },
    {
      "epoch": 1.136653895274585,
      "grad_norm": 1.0294743776321411,
      "learning_rate": 3.8810844103429463e-05,
      "loss": 0.4769,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 890
    },
    {
      "epoch": 1.1494252873563218,
      "grad_norm": 1.0869908332824707,
      "learning_rate": 3.8525175655594035e-05,
      "loss": 0.4913,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 900
    },
    {
      "epoch": 1.1621966794380587,
      "grad_norm": 1.0381580591201782,
      "learning_rate": 3.823767391838903e-05,
      "loss": 0.4817,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 910
    },
    {
      "epoch": 1.1749680715197957,
      "grad_norm": 0.9677332043647766,
      "learning_rate": 3.794842637040571e-05,
      "loss": 0.51,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 920
    },
    {
      "epoch": 1.1877394636015326,
      "grad_norm": 0.9293245673179626,
      "learning_rate": 3.765752102143586e-05,
      "loss": 0.5144,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 930
    },
    {
      "epoch": 1.2005108556832695,
      "grad_norm": 1.0104625225067139,
      "learning_rate": 3.736504638569289e-05,
      "loss": 0.4859,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 940
    },
    {
      "epoch": 1.2132822477650063,
      "grad_norm": 1.0003565549850464,
      "learning_rate": 3.70710914548795e-05,
      "loss": 0.4943,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 950
    },
    {
      "epoch": 1.2260536398467432,
      "grad_norm": 1.0630241632461548,
      "learning_rate": 3.6775745671110015e-05,
      "loss": 0.5145,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 960
    },
    {
      "epoch": 1.2388250319284801,
      "grad_norm": 0.9602718353271484,
      "learning_rate": 3.647909889969568e-05,
      "loss": 0.5117,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 970
    },
    {
      "epoch": 1.251596424010217,
      "grad_norm": 0.9984440207481384,
      "learning_rate": 3.618124140180128e-05,
      "loss": 0.5008,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 980
    },
    {
      "epoch": 1.264367816091954,
      "grad_norm": 1.0400162935256958,
      "learning_rate": 3.588226380698114e-05,
      "loss": 0.5055,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 990
    },
    {
      "epoch": 1.277139208173691,
      "grad_norm": 0.9871404767036438,
      "learning_rate": 3.558225708560322e-05,
      "loss": 0.5049,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1000
    },
    {
      "epoch": 1.289910600255428,
      "grad_norm": 0.9722399711608887,
      "learning_rate": 3.5281312521169374e-05,
      "loss": 0.5105,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1010
    },
    {
      "epoch": 1.3026819923371646,
      "grad_norm": 0.9392147064208984,
      "learning_rate": 3.4979521682540386e-05,
      "loss": 0.5036,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1020
    },
    {
      "epoch": 1.3154533844189016,
      "grad_norm": 0.9448984265327454,
      "learning_rate": 3.467697639607416e-05,
      "loss": 0.5043,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1030
    },
    {
      "epoch": 1.3282247765006385,
      "grad_norm": 1.0200536251068115,
      "learning_rate": 3.437376871768563e-05,
      "loss": 0.5059,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1040
    },
    {
      "epoch": 1.3409961685823755,
      "grad_norm": 1.0225882530212402,
      "learning_rate": 3.40699909048367e-05,
      "loss": 0.4928,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1050
    },
    {
      "epoch": 1.3537675606641124,
      "grad_norm": 1.0365824699401855,
      "learning_rate": 3.376573538846497e-05,
      "loss": 0.4946,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1060
    },
    {
      "epoch": 1.3665389527458494,
      "grad_norm": 0.8959736824035645,
      "learning_rate": 3.346109474485961e-05,
      "loss": 0.5091,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1070
    },
    {
      "epoch": 1.3793103448275863,
      "grad_norm": 0.9978928565979004,
      "learning_rate": 3.3156161667493045e-05,
      "loss": 0.5095,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1080
    },
    {
      "epoch": 1.392081736909323,
      "grad_norm": 0.9801906943321228,
      "learning_rate": 3.2851028938816954e-05,
      "loss": 0.5075,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1090
    },
    {
      "epoch": 1.40485312899106,
      "grad_norm": 1.1183043718338013,
      "learning_rate": 3.254578940203124e-05,
      "loss": 0.4852,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1100
    },
    {
      "epoch": 1.417624521072797,
      "grad_norm": 1.033026933670044,
      "learning_rate": 3.224053593283445e-05,
      "loss": 0.4955,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1110
    },
    {
      "epoch": 1.4303959131545338,
      "grad_norm": 1.0687352418899536,
      "learning_rate": 3.193536141116439e-05,
      "loss": 0.5326,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1120
    },
    {
      "epoch": 1.4431673052362708,
      "grad_norm": 0.8465182781219482,
      "learning_rate": 3.163035869293738e-05,
      "loss": 0.4994,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1130
    },
    {
      "epoch": 1.4559386973180077,
      "grad_norm": 1.1057614088058472,
      "learning_rate": 3.1325620581794824e-05,
      "loss": 0.5226,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1140
    },
    {
      "epoch": 1.4687100893997447,
      "grad_norm": 1.046042561531067,
      "learning_rate": 3.102123980086575e-05,
      "loss": 0.5076,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1150
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 1.0644861459732056,
      "learning_rate": 3.071730896455376e-05,
      "loss": 0.5354,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1160
    },
    {
      "epoch": 1.4942528735632183,
      "grad_norm": 0.8965176939964294,
      "learning_rate": 3.041392055035714e-05,
      "loss": 0.5115,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1170
    },
    {
      "epoch": 1.5070242656449553,
      "grad_norm": 0.9628166556358337,
      "learning_rate": 3.0111166870730545e-05,
      "loss": 0.5057,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1180
    },
    {
      "epoch": 1.5197956577266922,
      "grad_norm": 0.9256663918495178,
      "learning_rate": 2.9809140044997008e-05,
      "loss": 0.5159,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1190
    },
    {
      "epoch": 1.5325670498084292,
      "grad_norm": 1.103409767150879,
      "learning_rate": 2.9507931971318607e-05,
      "loss": 0.5146,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1200
    },
    {
      "epoch": 1.545338441890166,
      "grad_norm": 0.9358534812927246,
      "learning_rate": 2.9207634298734478e-05,
      "loss": 0.4867,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1210
    },
    {
      "epoch": 1.558109833971903,
      "grad_norm": 0.8767088055610657,
      "learning_rate": 2.8908338399274638e-05,
      "loss": 0.518,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1220
    },
    {
      "epoch": 1.5708812260536398,
      "grad_norm": 0.952031135559082,
      "learning_rate": 2.8610135340158082e-05,
      "loss": 0.4894,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1230
    },
    {
      "epoch": 1.5836526181353767,
      "grad_norm": 1.12105131149292,
      "learning_rate": 2.831311585608359e-05,
      "loss": 0.4952,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1240
    },
    {
      "epoch": 1.5964240102171137,
      "grad_norm": 0.979909360408783,
      "learning_rate": 2.8017370321621773e-05,
      "loss": 0.5033,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1250
    },
    {
      "epoch": 1.6091954022988506,
      "grad_norm": 0.9840870499610901,
      "learning_rate": 2.7722988723716642e-05,
      "loss": 0.5034,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1260
    },
    {
      "epoch": 1.6219667943805876,
      "grad_norm": 0.9386574625968933,
      "learning_rate": 2.7430060634305177e-05,
      "loss": 0.5204,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1270
    },
    {
      "epoch": 1.6347381864623243,
      "grad_norm": 0.9949301481246948,
      "learning_rate": 2.7138675183063095e-05,
      "loss": 0.5119,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1280
    },
    {
      "epoch": 1.6475095785440614,
      "grad_norm": 0.9763074517250061,
      "learning_rate": 2.6848921030285257e-05,
      "loss": 0.5174,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1290
    },
    {
      "epoch": 1.6602809706257982,
      "grad_norm": 1.006456732749939,
      "learning_rate": 2.6560886339908864e-05,
      "loss": 0.5217,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1300
    },
    {
      "epoch": 1.673052362707535,
      "grad_norm": 1.2566996812820435,
      "learning_rate": 2.6274658752687666e-05,
      "loss": 0.4763,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1310
    },
    {
      "epoch": 1.685823754789272,
      "grad_norm": 0.8630032539367676,
      "learning_rate": 2.599032535952537e-05,
      "loss": 0.4893,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1320
    },
    {
      "epoch": 1.698595146871009,
      "grad_norm": 0.9676541686058044,
      "learning_rate": 2.570797267497641e-05,
      "loss": 0.4789,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1330
    },
    {
      "epoch": 1.711366538952746,
      "grad_norm": 0.864613950252533,
      "learning_rate": 2.5427686610921963e-05,
      "loss": 0.5036,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1340
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 1.0790579319000244,
      "learning_rate": 2.5149552450429454e-05,
      "loss": 0.5045,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1350
    },
    {
      "epoch": 1.7369093231162198,
      "grad_norm": 1.2631795406341553,
      "learning_rate": 2.487365482180334e-05,
      "loss": 0.4956,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1360
    },
    {
      "epoch": 1.7496807151979565,
      "grad_norm": 0.9293349981307983,
      "learning_rate": 2.4600077672835098e-05,
      "loss": 0.5119,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1370
    },
    {
      "epoch": 1.7624521072796935,
      "grad_norm": 1.0236011743545532,
      "learning_rate": 2.4328904245260363e-05,
      "loss": 0.5174,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1380
    },
    {
      "epoch": 1.7752234993614304,
      "grad_norm": 0.9421340823173523,
      "learning_rate": 2.4060217049430776e-05,
      "loss": 0.4974,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1390
    },
    {
      "epoch": 1.7879948914431671,
      "grad_norm": 0.981686532497406,
      "learning_rate": 2.3794097839208465e-05,
      "loss": 0.4943,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1400
    },
    {
      "epoch": 1.8007662835249043,
      "grad_norm": 0.9267678260803223,
      "learning_rate": 2.353062758709067e-05,
      "loss": 0.5007,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1410
    },
    {
      "epoch": 1.813537675606641,
      "grad_norm": 0.9641256928443909,
      "learning_rate": 2.3269886459572126e-05,
      "loss": 0.5044,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1420
    },
    {
      "epoch": 1.8263090676883782,
      "grad_norm": 0.8920437693595886,
      "learning_rate": 2.3011953792752644e-05,
      "loss": 0.4932,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1430
    },
    {
      "epoch": 1.839080459770115,
      "grad_norm": 1.0809270143508911,
      "learning_rate": 2.2756908068197452e-05,
      "loss": 0.4959,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1440
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 0.9170002341270447,
      "learning_rate": 2.2504826889057443e-05,
      "loss": 0.4685,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1450
    },
    {
      "epoch": 1.8646232439335888,
      "grad_norm": 1.005517840385437,
      "learning_rate": 2.225578695645677e-05,
      "loss": 0.4891,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1460
    },
    {
      "epoch": 1.8773946360153255,
      "grad_norm": 0.9531996250152588,
      "learning_rate": 2.2009864046154872e-05,
      "loss": 0.4886,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1470
    },
    {
      "epoch": 1.8901660280970627,
      "grad_norm": 1.0945998430252075,
      "learning_rate": 2.176713298549006e-05,
      "loss": 0.4925,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1480
    },
    {
      "epoch": 1.9029374201787994,
      "grad_norm": 0.9576834440231323,
      "learning_rate": 2.1527667630611696e-05,
      "loss": 0.4681,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1490
    },
    {
      "epoch": 1.9157088122605364,
      "grad_norm": 0.9448633790016174,
      "learning_rate": 2.129154084400789e-05,
      "loss": 0.4659,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1500
    },
    {
      "epoch": 1.9284802043422733,
      "grad_norm": 1.0053128004074097,
      "learning_rate": 2.1058824472335476e-05,
      "loss": 0.495,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1510
    },
    {
      "epoch": 1.9412515964240102,
      "grad_norm": 0.9085705280303955,
      "learning_rate": 2.082958932455916e-05,
      "loss": 0.5215,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1520
    },
    {
      "epoch": 1.9540229885057472,
      "grad_norm": 0.8217594027519226,
      "learning_rate": 2.0603905150406348e-05,
      "loss": 0.5074,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1530
    },
    {
      "epoch": 1.966794380587484,
      "grad_norm": 0.8788007497787476,
      "learning_rate": 2.038184061914429e-05,
      "loss": 0.4869,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1540
    },
    {
      "epoch": 1.979565772669221,
      "grad_norm": 0.9042425751686096,
      "learning_rate": 2.0163463298686e-05,
      "loss": 0.4743,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1550
    },
    {
      "epoch": 1.9923371647509578,
      "grad_norm": 0.999984860420227,
      "learning_rate": 1.9948839635031224e-05,
      "loss": 0.5016,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1560
    },
    {
      "epoch": 2.005108556832695,
      "grad_norm": 0.9308561086654663,
      "learning_rate": 1.973803493204886e-05,
      "loss": 0.4209,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1570
    },
    {
      "epoch": 2.0178799489144317,
      "grad_norm": 1.1101335287094116,
      "learning_rate": 1.9531113331606815e-05,
      "loss": 0.3488,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1580
    },
    {
      "epoch": 2.0306513409961684,
      "grad_norm": 1.0135533809661865,
      "learning_rate": 1.932813779405547e-05,
      "loss": 0.3186,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1590
    },
    {
      "epoch": 2.0434227330779056,
      "grad_norm": 0.9131258726119995,
      "learning_rate": 1.9129170079070662e-05,
      "loss": 0.3323,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1600
    },
    {
      "epoch": 2.0561941251596423,
      "grad_norm": 1.0538150072097778,
      "learning_rate": 1.8934270726861937e-05,
      "loss": 0.3353,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1610
    },
    {
      "epoch": 2.0689655172413794,
      "grad_norm": 1.0463685989379883,
      "learning_rate": 1.8743499039751912e-05,
      "loss": 0.3349,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1620
    },
    {
      "epoch": 2.081736909323116,
      "grad_norm": 0.975685179233551,
      "learning_rate": 1.855691306413225e-05,
      "loss": 0.3113,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1630
    },
    {
      "epoch": 2.0945083014048533,
      "grad_norm": 0.9912129044532776,
      "learning_rate": 1.837456957280183e-05,
      "loss": 0.3332,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1640
    },
    {
      "epoch": 2.10727969348659,
      "grad_norm": 0.9777137041091919,
      "learning_rate": 1.819652404769235e-05,
      "loss": 0.3352,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1650
    },
    {
      "epoch": 2.1200510855683268,
      "grad_norm": 0.9486461281776428,
      "learning_rate": 1.8022830662986804e-05,
      "loss": 0.3377,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1660
    },
    {
      "epoch": 2.132822477650064,
      "grad_norm": 1.0125535726547241,
      "learning_rate": 1.7853542268635784e-05,
      "loss": 0.3617,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1670
    },
    {
      "epoch": 2.1455938697318007,
      "grad_norm": 1.044647455215454,
      "learning_rate": 1.7688710374276774e-05,
      "loss": 0.3523,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1680
    },
    {
      "epoch": 2.158365261813538,
      "grad_norm": 0.9924995303153992,
      "learning_rate": 1.7528385133561214e-05,
      "loss": 0.3313,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1690
    },
    {
      "epoch": 2.1711366538952745,
      "grad_norm": 1.0039716958999634,
      "learning_rate": 1.7372615328894174e-05,
      "loss": 0.3473,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1700
    },
    {
      "epoch": 2.1839080459770113,
      "grad_norm": 0.9947459697723389,
      "learning_rate": 1.7221448356591268e-05,
      "loss": 0.3357,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1710
    },
    {
      "epoch": 2.1966794380587484,
      "grad_norm": 0.9790370464324951,
      "learning_rate": 1.7074930212457307e-05,
      "loss": 0.3379,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1720
    },
    {
      "epoch": 2.209450830140485,
      "grad_norm": 0.9828362464904785,
      "learning_rate": 1.6933105477791058e-05,
      "loss": 0.3468,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1730
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.9688906073570251,
      "learning_rate": 1.679601730582046e-05,
      "loss": 0.3417,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1740
    },
    {
      "epoch": 2.234993614303959,
      "grad_norm": 1.0792521238327026,
      "learning_rate": 1.666370740857231e-05,
      "loss": 0.331,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1750
    },
    {
      "epoch": 2.247765006385696,
      "grad_norm": 1.001132607460022,
      "learning_rate": 1.653621604418046e-05,
      "loss": 0.3514,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1760
    },
    {
      "epoch": 2.260536398467433,
      "grad_norm": 0.9183688759803772,
      "learning_rate": 1.641358200463646e-05,
      "loss": 0.3307,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1770
    },
    {
      "epoch": 2.27330779054917,
      "grad_norm": 1.0465162992477417,
      "learning_rate": 1.6295842603986227e-05,
      "loss": 0.3456,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1780
    },
    {
      "epoch": 2.286079182630907,
      "grad_norm": 1.0534127950668335,
      "learning_rate": 1.6183033666976434e-05,
      "loss": 0.3327,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1790
    },
    {
      "epoch": 2.2988505747126435,
      "grad_norm": 1.023949384689331,
      "learning_rate": 1.6075189518154097e-05,
      "loss": 0.3432,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1800
    },
    {
      "epoch": 2.3116219667943807,
      "grad_norm": 0.9274671077728271,
      "learning_rate": 1.597234297142256e-05,
      "loss": 0.3566,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1810
    },
    {
      "epoch": 2.3243933588761174,
      "grad_norm": 0.91460782289505,
      "learning_rate": 1.587452532005717e-05,
      "loss": 0.3333,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1820
    },
    {
      "epoch": 2.3371647509578546,
      "grad_norm": 0.9292669892311096,
      "learning_rate": 1.5781766327183632e-05,
      "loss": 0.3485,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1830
    },
    {
      "epoch": 2.3499361430395913,
      "grad_norm": 0.9177665710449219,
      "learning_rate": 1.569409421672192e-05,
      "loss": 0.3294,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1840
    },
    {
      "epoch": 2.362707535121328,
      "grad_norm": 1.0396416187286377,
      "learning_rate": 1.561153566479856e-05,
      "loss": 0.3509,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1850
    },
    {
      "epoch": 2.375478927203065,
      "grad_norm": 1.0500093698501587,
      "learning_rate": 1.5534115791629842e-05,
      "loss": 0.3487,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1860
    },
    {
      "epoch": 2.388250319284802,
      "grad_norm": 0.926463782787323,
      "learning_rate": 1.5461858153878433e-05,
      "loss": 0.3496,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1870
    },
    {
      "epoch": 2.401021711366539,
      "grad_norm": 1.0274465084075928,
      "learning_rate": 1.539478473748578e-05,
      "loss": 0.3384,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1880
    },
    {
      "epoch": 2.413793103448276,
      "grad_norm": 1.0295958518981934,
      "learning_rate": 1.53329159509824e-05,
      "loss": 0.3485,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1890
    },
    {
      "epoch": 2.4265644955300125,
      "grad_norm": 1.017041563987732,
      "learning_rate": 1.527627061927816e-05,
      "loss": 0.3538,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1900
    },
    {
      "epoch": 2.4393358876117497,
      "grad_norm": 0.9540072679519653,
      "learning_rate": 1.5224865977934372e-05,
      "loss": 0.3437,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1910
    },
    {
      "epoch": 2.4521072796934864,
      "grad_norm": 0.9918265342712402,
      "learning_rate": 1.5178717667919515e-05,
      "loss": 0.3605,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1920
    },
    {
      "epoch": 2.4648786717752236,
      "grad_norm": 1.0125731229782104,
      "learning_rate": 1.5137839730850137e-05,
      "loss": 0.3665,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1930
    },
    {
      "epoch": 2.4776500638569603,
      "grad_norm": 0.9846736192703247,
      "learning_rate": 1.5102244604718386e-05,
      "loss": 0.3505,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1940
    },
    {
      "epoch": 2.4904214559386975,
      "grad_norm": 1.0192700624465942,
      "learning_rate": 1.5071943120107471e-05,
      "loss": 0.3386,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1950
    },
    {
      "epoch": 2.503192848020434,
      "grad_norm": 0.9997414350509644,
      "learning_rate": 1.5046944496896256e-05,
      "loss": 0.3487,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1960
    },
    {
      "epoch": 2.5159642401021713,
      "grad_norm": 0.9771548509597778,
      "learning_rate": 1.5027256341453883e-05,
      "loss": 0.3345,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1970
    },
    {
      "epoch": 2.528735632183908,
      "grad_norm": 0.9269956350326538,
      "learning_rate": 1.5012884644325379e-05,
      "loss": 0.3341,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1980
    },
    {
      "epoch": 2.541507024265645,
      "grad_norm": 1.0031806230545044,
      "learning_rate": 1.5003833778408907e-05,
      "loss": 0.3309,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 1990
    },
    {
      "epoch": 2.554278416347382,
      "grad_norm": 1.0676020383834839,
      "learning_rate": 1.5000106497625225e-05,
      "loss": 0.3431,
      "memory/device_memory_reserved(gib)": 33.64,
      "memory/max_memory_active(gib)": 31.61,
      "memory/max_memory_allocated(gib)": 31.61,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 2001,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.477704900591485e+18,
  "train_batch_size": 28,
  "trial_name": null,
  "trial_params": null
}