training_metrics.json

{
  "activation_checkpointing": true,
  "depth": 8,
  "eval_batch_size": 4,
  "img_size": 512,
  "mfu_percent": 9.28668914165394,
  "num_params": 77575312,
  "num_steps": 214,
  "patch_size": 14,
  "peak_vram_mb": 3816.359375,
  "smoke_test": false,
  "tokenizer_dataset": "multivision",
  "total_seconds": 4153.214341402054,
  "total_tokens": 112197632,
  "train_batch_size": 4,
  "train_dataset": "multivision",
  "training_seconds": 3605.0291242599487,
  "use_moe": false,
  "use_vision": true,
  "val_bpb": 1.2905162668945176,
  "vision_dim": 128,
  "vision_heads": 4,
  "vision_layers": 4
}