1. Download the dataset from https://demuc.de/colmap/datasets/
2. Train it with LFS
3. Settings were used:
* using bilateral-grid
* 60k steps+15k steps for sparsify
* 7,5 M splats (before sparsify), 3 M splats (final)
* means_lr = 1.6e-05 and opacity_reg = 0.001
* resize_factor = 4 (I resize it with cv2.resize(img, (width/4, height/4), interpolation=cv2.INTER_AREA))