{
  "history": [
    {
      "step": 1,
      "mean_question_loss": 1.303297832608223,
      "seconds_per_step": 1.9828559500165284,
      "questions_per_second": 32.27667647741457,
      "tokens_per_second": 11061.317893423964,
      "peak_gpu_memory_gib": 33.7561559677124
    },
    {
      "step": 50,
      "mean_question_loss": 1.1774762195287918,
      "seconds_per_step": 0.9613677148371033,
      "questions_per_second": 66.57182159569851,
      "tokens_per_second": 23273.15644074954,
      "peak_gpu_memory_gib": 33.77996015548706
    },
    {
      "step": 100,
      "mean_question_loss": 1.0275238631665706,
      "seconds_per_step": 0.936516181200277,
      "questions_per_second": 68.33838142334605,
      "tokens_per_second": 23841.74502076761,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 150,
      "mean_question_loss": 1.0179613350331784,
      "seconds_per_step": 0.9423550958000123,
      "questions_per_second": 67.91495083460784,
      "tokens_per_second": 23626.656341363956,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 200,
      "mean_question_loss": 1.0347131621092558,
      "seconds_per_step": 0.9440174014202785,
      "questions_per_second": 67.79536044961853,
      "tokens_per_second": 23541.87535797961,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 250,
      "mean_question_loss": 1.0159618432074786,
      "seconds_per_step": 0.94187130271981,
      "questions_per_second": 67.9498354129586,
      "tokens_per_second": 23729.16547670704,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 300,
      "mean_question_loss": 1.0153298786282539,
      "seconds_per_step": 0.9525986466201721,
      "questions_per_second": 67.18464300476653,
      "tokens_per_second": 23441.036872377124,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 350,
      "mean_question_loss": 0.9680072110146284,
      "seconds_per_step": 0.9382304856198607,
      "questions_per_second": 68.2135157415154,
      "tokens_per_second": 23672.989036724855,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 400,
      "mean_question_loss": 0.9385269486159086,
      "seconds_per_step": 0.9443311922001886,
      "questions_per_second": 67.77283280338013,
      "tokens_per_second": 23629.99357038028,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 450,
      "mean_question_loss": 0.9065450581163168,
      "seconds_per_step": 0.9466447888000402,
      "questions_per_second": 67.60719623368541,
      "tokens_per_second": 23501.867081739598,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 500,
      "mean_question_loss": 0.9177573589235544,
      "seconds_per_step": 0.9342501087801065,
      "questions_per_second": 68.5041397357371,
      "tokens_per_second": 23866.671023581464,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 550,
      "mean_question_loss": 0.8672973342984914,
      "seconds_per_step": 0.9519798879802692,
      "questions_per_second": 67.22831102638428,
      "tokens_per_second": 23354.65305592759,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 600,
      "mean_question_loss": 0.8056097267568112,
      "seconds_per_step": 0.9394654304598226,
      "questions_per_second": 68.12384780212203,
      "tokens_per_second": 23848.711483757124,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 650,
      "mean_question_loss": 0.7258428598567843,
      "seconds_per_step": 0.9375208237400511,
      "questions_per_second": 68.26515036187126,
      "tokens_per_second": 23777.19986108901,
      "peak_gpu_memory_gib": 33.77999830245972
    },
    {
      "step": 700,
      "mean_question_loss": 0.7136679108440876,
      "seconds_per_step": 0.9387312200397718,
      "questions_per_second": 68.1771295486353,
      "tokens_per_second": 23724.85278486468,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 750,
      "mean_question_loss": 0.6691893465071916,
      "seconds_per_step": 0.9545819705398754,
      "questions_per_second": 67.04505424903849,
      "tokens_per_second": 23300.79625055193,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 800,
      "mean_question_loss": 0.6290645541995764,
      "seconds_per_step": 0.93820280096028,
      "questions_per_second": 68.21552859839471,
      "tokens_per_second": 23736.104792275928,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 850,
      "mean_question_loss": 0.6105163471773267,
      "seconds_per_step": 0.9443530672800261,
      "questions_per_second": 67.77126290735313,
      "tokens_per_second": 23599.055027362607,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 900,
      "mean_question_loss": 0.5953477524220944,
      "seconds_per_step": 0.9439651129196863,
      "questions_per_second": 67.79911579787927,
      "tokens_per_second": 23561.93009210537,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 950,
      "mean_question_loss": 0.593899949491024,
      "seconds_per_step": 0.976987686319626,
      "questions_per_second": 65.507478647036,
      "tokens_per_second": 22949.48064746105,
      "peak_gpu_memory_gib": 33.78000593185425
    },
    {
      "step": 1000,
      "mean_question_loss": 0.5632530002482236,
      "seconds_per_step": 0.954888053460163,
      "questions_per_second": 67.02356340943584,
      "tokens_per_second": 23455.796644265385,
      "peak_gpu_memory_gib": 33.78000593185425
    }
  ],
  "steps": 1000,
  "world_size": 4,
  "effective_batch": 64,
  "questions_per_device": 8,
  "accumulation": 2,
  "encoding_microbatch": 32,
  "gradient_checkpointing": false,
  "pretokenized": true,
  "mixed_precision": "bf16",
  "compute_dtype": "torch.bfloat16",
  "attention_implementation": "sdpa",
  "trainable_parameters": 2129920
}