{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.597938144329897,
  "eval_steps": 500,
  "global_step": 78,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020618556701030927,
      "grad_norm": 0.3655090928077698,
      "learning_rate": 0.0,
      "loss": 0.9385213255882263,
      "memory/device_reserved (GiB)": 108.13,
      "memory/max_active (GiB)": 104.61,
      "memory/max_allocated (GiB)": 104.61,
      "ppl": 2.5562,
      "step": 1,
      "tokens/total": 262144,
      "tokens/train_per_sec_per_gpu": 217.44,
      "tokens/trainable": 215045
    },
    {
      "epoch": 0.041237113402061855,
      "grad_norm": 0.3760335445404053,
      "learning_rate": 2.499999936844688e-06,
      "loss": 0.9638054966926575,
      "memory/device_reserved (GiB)": 108.37,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.62165,
      "step": 2,
      "tokens/total": 524288,
      "tokens/train_per_sec_per_gpu": 384.38,
      "tokens/trainable": 441785
    },
    {
      "epoch": 0.061855670103092786,
      "grad_norm": 0.33114928007125854,
      "learning_rate": 4.999999873689376e-06,
      "loss": 0.9079572558403015,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.47925,
      "step": 3,
      "tokens/total": 786432,
      "tokens/train_per_sec_per_gpu": 396.12,
      "tokens/trainable": 656551
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 0.283974826335907,
      "learning_rate": 7.499999810534064e-06,
      "loss": 0.9048001766204834,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 104.63,
      "memory/max_allocated (GiB)": 104.63,
      "ppl": 2.47144,
      "step": 4,
      "tokens/total": 1048576,
      "tokens/train_per_sec_per_gpu": 440.85,
      "tokens/trainable": 879207
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 0.1975884884595871,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.8495224714279175,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.33853,
      "step": 5,
      "tokens/total": 1310720,
      "tokens/train_per_sec_per_gpu": 459.2,
      "tokens/trainable": 1107620
    },
    {
      "epoch": 0.12371134020618557,
      "grad_norm": 0.1656099259853363,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.8573672771453857,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 104.01,
      "memory/max_allocated (GiB)": 104.01,
      "ppl": 2.35695,
      "step": 6,
      "tokens/total": 1572864,
      "tokens/train_per_sec_per_gpu": 449.42,
      "tokens/trainable": 1335294
    },
    {
      "epoch": 0.14432989690721648,
      "grad_norm": 0.19185799360275269,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.828778862953186,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.29052,
      "step": 7,
      "tokens/total": 1835008,
      "tokens/train_per_sec_per_gpu": 418.68,
      "tokens/trainable": 1551508
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 0.19206257164478302,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.8039666414260864,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.23439,
      "step": 8,
      "tokens/total": 2097152,
      "tokens/train_per_sec_per_gpu": 447.93,
      "tokens/trainable": 1780305
    },
    {
      "epoch": 0.18556701030927836,
      "grad_norm": 0.17175284028053284,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.835797905921936,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.30665,
      "step": 9,
      "tokens/total": 2359296,
      "tokens/train_per_sec_per_gpu": 375.13,
      "tokens/trainable": 1986961
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 0.15307827293872833,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.857429563999176,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.35709,
      "step": 10,
      "tokens/total": 2621440,
      "tokens/train_per_sec_per_gpu": 438.82,
      "tokens/trainable": 2206017
    },
    {
      "epoch": 0.2268041237113402,
      "grad_norm": 0.12776634097099304,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.827063798904419,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.28659,
      "step": 11,
      "tokens/total": 2883584,
      "tokens/train_per_sec_per_gpu": 400.94,
      "tokens/trainable": 2423186
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 0.11723587661981583,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7730377912521362,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.16634,
      "step": 12,
      "tokens/total": 3145728,
      "tokens/train_per_sec_per_gpu": 377.71,
      "tokens/trainable": 2641585
    },
    {
      "epoch": 0.26804123711340205,
      "grad_norm": 0.10886862874031067,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7753809690475464,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.17142,
      "step": 13,
      "tokens/total": 3407872,
      "tokens/train_per_sec_per_gpu": 437.36,
      "tokens/trainable": 2866591
    },
    {
      "epoch": 0.28865979381443296,
      "grad_norm": 0.10167527943849564,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7707717418670654,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.16143,
      "step": 14,
      "tokens/total": 3670016,
      "tokens/train_per_sec_per_gpu": 424.92,
      "tokens/trainable": 3093104
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 0.0818706750869751,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7532338500022888,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.12386,
      "step": 15,
      "tokens/total": 3932160,
      "tokens/train_per_sec_per_gpu": 422.14,
      "tokens/trainable": 3316386
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 0.0788501426577568,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7692117094993591,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.15806,
      "step": 16,
      "tokens/total": 4194304,
      "tokens/train_per_sec_per_gpu": 446.3,
      "tokens/trainable": 3541549
    },
    {
      "epoch": 0.35051546391752575,
      "grad_norm": 0.07898039370775223,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7503528594970703,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.11775,
      "step": 17,
      "tokens/total": 4456448,
      "tokens/train_per_sec_per_gpu": 439.26,
      "tokens/trainable": 3764126
    },
    {
      "epoch": 0.3711340206185567,
      "grad_norm": 0.08236191421747208,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7804020643234253,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.18235,
      "step": 18,
      "tokens/total": 4718592,
      "tokens/train_per_sec_per_gpu": 422.06,
      "tokens/trainable": 3982168
    },
    {
      "epoch": 0.3917525773195876,
      "grad_norm": 0.07432285696268082,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7135086059570312,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.04114,
      "step": 19,
      "tokens/total": 4980736,
      "tokens/train_per_sec_per_gpu": 425.85,
      "tokens/trainable": 4202105
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 0.07066275924444199,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7594545483589172,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.13711,
      "step": 20,
      "tokens/total": 5242880,
      "tokens/train_per_sec_per_gpu": 423.14,
      "tokens/trainable": 4425697
    },
    {
      "epoch": 0.4329896907216495,
      "grad_norm": 0.06695688515901566,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7534548044204712,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.12433,
      "step": 21,
      "tokens/total": 5505024,
      "tokens/train_per_sec_per_gpu": 444.63,
      "tokens/trainable": 4638352
    },
    {
      "epoch": 0.4536082474226804,
      "grad_norm": 0.06697328388690948,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7117219567298889,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.0375,
      "step": 22,
      "tokens/total": 5767168,
      "tokens/train_per_sec_per_gpu": 431.44,
      "tokens/trainable": 4859204
    },
    {
      "epoch": 0.4742268041237113,
      "grad_norm": 0.06299509853124619,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7019100785255432,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.0176,
      "step": 23,
      "tokens/total": 6029312,
      "tokens/train_per_sec_per_gpu": 432.19,
      "tokens/trainable": 5082925
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 0.06494089215993881,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.69496750831604,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.00364,
      "step": 24,
      "tokens/total": 6291456,
      "tokens/train_per_sec_per_gpu": 427.22,
      "tokens/trainable": 5305013
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 0.06321357190608978,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6761184930801392,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.96623,
      "step": 25,
      "tokens/total": 6553600,
      "tokens/train_per_sec_per_gpu": 439.28,
      "tokens/trainable": 5532999
    },
    {
      "epoch": 0.5360824742268041,
      "grad_norm": 0.06567544490098953,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7003852725028992,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.01453,
      "step": 26,
      "tokens/total": 6815744,
      "tokens/train_per_sec_per_gpu": 454.59,
      "tokens/trainable": 5763467
    },
    {
      "epoch": 0.5567010309278351,
      "grad_norm": 0.09651881456375122,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6703989505767822,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.95502,
      "step": 27,
      "tokens/total": 7077888,
      "tokens/train_per_sec_per_gpu": 431.21,
      "tokens/trainable": 5979795
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 0.06766466051340103,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7193452715873718,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.05309,
      "step": 28,
      "tokens/total": 7340032,
      "tokens/train_per_sec_per_gpu": 398.38,
      "tokens/trainable": 6200496
    },
    {
      "epoch": 0.5979381443298969,
      "grad_norm": 0.06165790185332298,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7149662971496582,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.04412,
      "step": 29,
      "tokens/total": 7602176,
      "tokens/train_per_sec_per_gpu": 408.69,
      "tokens/trainable": 6412932
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 0.06306911259889603,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7180032134056091,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.05034,
      "step": 30,
      "tokens/total": 7864320,
      "tokens/train_per_sec_per_gpu": 402.32,
      "tokens/trainable": 6628700
    },
    {
      "epoch": 0.6391752577319587,
      "grad_norm": 0.0602366179227829,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7157541513442993,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.04573,
      "step": 31,
      "tokens/total": 8126464,
      "tokens/train_per_sec_per_gpu": 416.38,
      "tokens/trainable": 6857590
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 0.06123146787285805,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7323997020721436,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.08007,
      "step": 32,
      "tokens/total": 8388608,
      "tokens/train_per_sec_per_gpu": 355.01,
      "tokens/trainable": 7073879
    },
    {
      "epoch": 0.6804123711340206,
      "grad_norm": 0.06059738248586655,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7036861181259155,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.02119,
      "step": 33,
      "tokens/total": 8650752,
      "tokens/train_per_sec_per_gpu": 419.84,
      "tokens/trainable": 7293612
    },
    {
      "epoch": 0.7010309278350515,
      "grad_norm": 0.0669514387845993,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7239381074905396,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.06254,
      "step": 34,
      "tokens/total": 8912896,
      "tokens/train_per_sec_per_gpu": 415.53,
      "tokens/trainable": 7511070
    },
    {
      "epoch": 0.7216494845360825,
      "grad_norm": 0.05671551451086998,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7183998823165894,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.05115,
      "step": 35,
      "tokens/total": 9175040,
      "tokens/train_per_sec_per_gpu": 449.92,
      "tokens/trainable": 7734018
    },
    {
      "epoch": 0.7422680412371134,
      "grad_norm": 0.06342059373855591,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7117175459861755,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.03749,
      "step": 36,
      "tokens/total": 9437184,
      "tokens/train_per_sec_per_gpu": 444.82,
      "tokens/trainable": 7964855
    },
    {
      "epoch": 0.7628865979381443,
      "grad_norm": 0.057359274476766586,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7321305871009827,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.07951,
      "step": 37,
      "tokens/total": 9699328,
      "tokens/train_per_sec_per_gpu": 415.77,
      "tokens/trainable": 8186023
    },
    {
      "epoch": 0.7835051546391752,
      "grad_norm": 0.05887288600206375,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7334283590316772,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.08221,
      "step": 38,
      "tokens/total": 9961472,
      "tokens/train_per_sec_per_gpu": 413.7,
      "tokens/trainable": 8411016
    },
    {
      "epoch": 0.8041237113402062,
      "grad_norm": 0.056856513023376465,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7016430497169495,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.01706,
      "step": 39,
      "tokens/total": 10223616,
      "tokens/train_per_sec_per_gpu": 434.75,
      "tokens/trainable": 8630090
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.05889676883816719,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7155553102493286,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.04532,
      "step": 40,
      "tokens/total": 10485760,
      "tokens/train_per_sec_per_gpu": 418.82,
      "tokens/trainable": 8847803
    },
    {
      "epoch": 0.845360824742268,
      "grad_norm": 0.05640722066164017,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7292789816856384,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.07358,
      "step": 41,
      "tokens/total": 10747904,
      "tokens/train_per_sec_per_gpu": 443.09,
      "tokens/trainable": 9073730
    },
    {
      "epoch": 0.865979381443299,
      "grad_norm": 0.058233924210071564,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6932827234268188,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.00027,
      "step": 42,
      "tokens/total": 11010048,
      "tokens/train_per_sec_per_gpu": 405.02,
      "tokens/trainable": 9293503
    },
    {
      "epoch": 0.8865979381443299,
      "grad_norm": 0.05681903660297394,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6836936473846436,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98118,
      "step": 43,
      "tokens/total": 11272192,
      "tokens/train_per_sec_per_gpu": 428.1,
      "tokens/trainable": 9513886
    },
    {
      "epoch": 0.9072164948453608,
      "grad_norm": 0.05938011407852173,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6787852644920349,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.97148,
      "step": 44,
      "tokens/total": 11534336,
      "tokens/train_per_sec_per_gpu": 402.73,
      "tokens/trainable": 9738138
    },
    {
      "epoch": 0.9278350515463918,
      "grad_norm": 0.27452051639556885,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6889349222183228,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.99159,
      "step": 45,
      "tokens/total": 11796480,
      "tokens/train_per_sec_per_gpu": 418.07,
      "tokens/trainable": 9964898
    },
    {
      "epoch": 0.9484536082474226,
      "grad_norm": 0.052989210933446884,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6635887622833252,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.94175,
      "step": 46,
      "tokens/total": 12058624,
      "tokens/train_per_sec_per_gpu": 457.13,
      "tokens/trainable": 10194034
    },
    {
      "epoch": 0.9690721649484536,
      "grad_norm": 0.05614595487713814,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7070361375808716,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.02797,
      "step": 47,
      "tokens/total": 12320768,
      "tokens/train_per_sec_per_gpu": 439.6,
      "tokens/trainable": 10429548
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 0.0553651824593544,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6467244625091553,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.90928,
      "step": 48,
      "tokens/total": 12582912,
      "tokens/train_per_sec_per_gpu": 438.08,
      "tokens/trainable": 10646710
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0849444791674614,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7548460960388184,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 104.0,
      "memory/max_allocated (GiB)": 104.0,
      "ppl": 2.12728,
      "step": 49,
      "tokens/total": 12697600,
      "tokens/train_per_sec_per_gpu": 383.2,
      "tokens/trainable": 10742066
    },
    {
      "epoch": 1.0206185567010309,
      "grad_norm": 0.05694010481238365,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.667358934879303,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.94908,
      "step": 50,
      "tokens/total": 12959744,
      "tokens/train_per_sec_per_gpu": 426.98,
      "tokens/trainable": 10967540
    },
    {
      "epoch": 1.041237113402062,
      "grad_norm": 0.0527355931699276,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6806523203849792,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.97517,
      "step": 51,
      "tokens/total": 13221888,
      "tokens/train_per_sec_per_gpu": 402.07,
      "tokens/trainable": 11192090
    },
    {
      "epoch": 1.0618556701030928,
      "grad_norm": 0.06241290643811226,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6800387501716614,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.97395,
      "step": 52,
      "tokens/total": 13484032,
      "tokens/train_per_sec_per_gpu": 445.75,
      "tokens/trainable": 11412434
    },
    {
      "epoch": 1.0824742268041236,
      "grad_norm": 0.05897771567106247,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6847533583641052,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98328,
      "step": 53,
      "tokens/total": 13746176,
      "tokens/train_per_sec_per_gpu": 443.35,
      "tokens/trainable": 11622868
    },
    {
      "epoch": 1.1030927835051547,
      "grad_norm": 0.05696937441825867,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6858651638031006,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98549,
      "step": 54,
      "tokens/total": 14008320,
      "tokens/train_per_sec_per_gpu": 437.03,
      "tokens/trainable": 11835833
    },
    {
      "epoch": 1.1237113402061856,
      "grad_norm": 0.05696974694728851,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6698517203330994,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.95395,
      "step": 55,
      "tokens/total": 14270464,
      "tokens/train_per_sec_per_gpu": 394.02,
      "tokens/trainable": 12052674
    },
    {
      "epoch": 1.1443298969072164,
      "grad_norm": 0.053235217928886414,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6868125200271606,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98737,
      "step": 56,
      "tokens/total": 14532608,
      "tokens/train_per_sec_per_gpu": 416.17,
      "tokens/trainable": 12280836
    },
    {
      "epoch": 1.1649484536082475,
      "grad_norm": 0.053780533373355865,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6769828200340271,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.96793,
      "step": 57,
      "tokens/total": 14794752,
      "tokens/train_per_sec_per_gpu": 427.65,
      "tokens/trainable": 12510063
    },
    {
      "epoch": 1.1855670103092784,
      "grad_norm": 0.05554712563753128,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6530709266662598,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.92143,
      "step": 58,
      "tokens/total": 15056896,
      "tokens/train_per_sec_per_gpu": 448.94,
      "tokens/trainable": 12734156
    },
    {
      "epoch": 1.2061855670103092,
      "grad_norm": 0.058325208723545074,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6604984998703003,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.93576,
      "step": 59,
      "tokens/total": 15319040,
      "tokens/train_per_sec_per_gpu": 426.43,
      "tokens/trainable": 12947517
    },
    {
      "epoch": 1.2268041237113403,
      "grad_norm": 0.059033673256635666,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6748796105384827,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.9638,
      "step": 60,
      "tokens/total": 15581184,
      "tokens/train_per_sec_per_gpu": 445.76,
      "tokens/trainable": 13169591
    },
    {
      "epoch": 1.2474226804123711,
      "grad_norm": 0.059492629021406174,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7338234186172485,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.08303,
      "step": 61,
      "tokens/total": 15843328,
      "tokens/train_per_sec_per_gpu": 437.44,
      "tokens/trainable": 13395658
    },
    {
      "epoch": 1.268041237113402,
      "grad_norm": 0.05484800785779953,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6759692430496216,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.96594,
      "step": 62,
      "tokens/total": 16105472,
      "tokens/train_per_sec_per_gpu": 435.01,
      "tokens/trainable": 13623398
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 0.05711665004491806,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6200935244560242,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.8591,
      "step": 63,
      "tokens/total": 16367616,
      "tokens/train_per_sec_per_gpu": 428.78,
      "tokens/trainable": 13847595
    },
    {
      "epoch": 1.309278350515464,
      "grad_norm": 0.05786803737282753,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6560422778129578,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.92715,
      "step": 64,
      "tokens/total": 16629760,
      "tokens/train_per_sec_per_gpu": 403.39,
      "tokens/trainable": 14062516
    },
    {
      "epoch": 1.3298969072164948,
      "grad_norm": 0.05644693970680237,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6657907962799072,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.94603,
      "step": 65,
      "tokens/total": 16891904,
      "tokens/train_per_sec_per_gpu": 418.07,
      "tokens/trainable": 14289902
    },
    {
      "epoch": 1.3505154639175259,
      "grad_norm": 0.06030877307057381,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.649398922920227,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.91439,
      "step": 66,
      "tokens/total": 17154048,
      "tokens/train_per_sec_per_gpu": 416.1,
      "tokens/trainable": 14496986
    },
    {
      "epoch": 1.3711340206185567,
      "grad_norm": 0.05981199070811272,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6490236520767212,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.91367,
      "step": 67,
      "tokens/total": 17416192,
      "tokens/train_per_sec_per_gpu": 421.32,
      "tokens/trainable": 14719990
    },
    {
      "epoch": 1.3917525773195876,
      "grad_norm": 0.05708547681570053,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6839650869369507,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98172,
      "step": 68,
      "tokens/total": 17678336,
      "tokens/train_per_sec_per_gpu": 415.26,
      "tokens/trainable": 14949006
    },
    {
      "epoch": 1.4123711340206184,
      "grad_norm": 0.06631551682949066,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.7076646089553833,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 2.02925,
      "step": 69,
      "tokens/total": 17940480,
      "tokens/train_per_sec_per_gpu": 402.9,
      "tokens/trainable": 15171377
    },
    {
      "epoch": 1.4329896907216495,
      "grad_norm": 0.057865116745233536,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6435024738311768,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.90313,
      "step": 70,
      "tokens/total": 18202624,
      "tokens/train_per_sec_per_gpu": 393.81,
      "tokens/trainable": 15391463
    },
    {
      "epoch": 1.4536082474226804,
      "grad_norm": 0.05982322245836258,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6845062971115112,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.98279,
      "step": 71,
      "tokens/total": 18464768,
      "tokens/train_per_sec_per_gpu": 421.18,
      "tokens/trainable": 15615119
    },
    {
      "epoch": 1.4742268041237114,
      "grad_norm": 0.059283241629600525,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6602850556373596,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.93534,
      "step": 72,
      "tokens/total": 18726912,
      "tokens/train_per_sec_per_gpu": 402.44,
      "tokens/trainable": 15843792
    },
    {
      "epoch": 1.4948453608247423,
      "grad_norm": 0.06889505684375763,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6537664532661438,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.92277,
      "step": 73,
      "tokens/total": 18989056,
      "tokens/train_per_sec_per_gpu": 424.9,
      "tokens/trainable": 16068606
    },
    {
      "epoch": 1.5154639175257731,
      "grad_norm": 0.058555394411087036,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6312741041183472,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.88,
      "step": 74,
      "tokens/total": 19251200,
      "tokens/train_per_sec_per_gpu": 421.77,
      "tokens/trainable": 16290030
    },
    {
      "epoch": 1.536082474226804,
      "grad_norm": 0.11190961301326752,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6814379692077637,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.97672,
      "step": 75,
      "tokens/total": 19513344,
      "tokens/train_per_sec_per_gpu": 449.37,
      "tokens/trainable": 16521988
    },
    {
      "epoch": 1.556701030927835,
      "grad_norm": 0.05849243700504303,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6415979862213135,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.89951,
      "step": 76,
      "tokens/total": 19775488,
      "tokens/train_per_sec_per_gpu": 428.61,
      "tokens/trainable": 16751706
    },
    {
      "epoch": 1.577319587628866,
      "grad_norm": 0.06144499406218529,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6616408824920654,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.93797,
      "step": 77,
      "tokens/total": 20037632,
      "tokens/train_per_sec_per_gpu": 391.36,
      "tokens/trainable": 16966762
    },
    {
      "epoch": 1.597938144329897,
      "grad_norm": 0.05783751979470253,
      "learning_rate": 9.999999747378752e-06,
      "loss": 0.6137974858283997,
      "memory/device_reserved (GiB)": 108.41,
      "memory/max_active (GiB)": 105.26,
      "memory/max_allocated (GiB)": 105.26,
      "ppl": 1.84743,
      "step": 78,
      "tokens/total": 20299776,
      "tokens/train_per_sec_per_gpu": 448.81,
      "tokens/trainable": 17187456
    }
  ],
  "logging_steps": 1,
  "max_steps": 96,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 6,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2185977290741514e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}