d8/meta_000200.json

{
  "step": 200,
  "val_bpb": 3.3393806087790194,
  "model_config": {
    "max_seq_len": 2048,
    "vocab_size": 65536,
    "n_layer": 8,
    "n_head": 4,
    "n_embd": 512,
    "n_hidden": 8192
  },
  "user_config": {
    "run": "bdh_train_21",
    "device_type": "",
    "depth": 8,
    "max_seq_len": 2048,
    "num_iterations": 5000,
    "target_flops": -1.0,
    "target_param_data_ratio": 20,
    "device_batch_size": 8,
    "total_batch_size": 524288,
    "embedding_lr": 0.002,
    "unembedding_lr": 0.004,
    "weight_decay": 0.1,
    "matrix_lr": 0.02,
    "grad_clip": 1.0,
    "warmup_ratio": 0.0,
    "warmdown_ratio": 0.2,
    "final_lr_frac": 0.0,
    "resume_from_step": -1,
    "eval_every": 1000,
    "eval_tokens": 10485760,
    "core_metric_every": 2500,
    "core_metric_max_per_task": 500,
    "sample_every": 100,
    "save_every": 100,
    "model_tag": ""
  },
  "device_batch_size": 8,
  "max_seq_len": 2048,
  "dataloader_state_dict": {
    "batches_yielded": 800
  },
  "loop_state": {
    "min_val_bpb": 3.3393806087790194,
    "smooth_train_loss": 5.6278228485674235,
    "total_training_time": 713.754873752594
  }
}