training_metrics.json

{
  "activation_checkpointing": true,
  "depth": 8,
  "eval_batch_size": 4,
  "img_size": 512,
  "mfu_percent": 9.23946110204506,
  "num_params": 77575312,
  "num_steps": 213,
  "patch_size": 14,
  "peak_vram_mb": 3815.390625,
  "smoke_test": false,
  "tokenizer_dataset": "multivision",
  "total_seconds": 4155.633095502853,
  "total_tokens": 111673344,
  "train_batch_size": 4,
  "train_dataset": "multivision",
  "training_seconds": 3605.694397687912,
  "use_moe": false,
  "use_vision": true,
  "val_bpb": 2.4157603828717646,
  "vision_dim": 128,
  "vision_heads": 4,
  "vision_layers": 4
}