ValueError: not enough values to unpack (expected 4, got 2)

[xjli360]

Wonderful work! But when I run: python train_3DGS.py --fp16 --workspace ./res_gaussion/colmap_doll \ --test --sample \ --bounding_box_path './data/3d_box/doll_glass_bbox.ply' \ --radius_list 1.3 --fovy 50 --phi_list -45 -30 -15 0 15 30 45 --theta_list 60 75 90

i got:

Namespace(test=True, fp16=True, save_video=False, eval_interval=50, workspace='./res_gaussion/colmap_doll', seed=0, sample=True, radius_list=[1.3], fovy=50.0, phi_list=[-45.0, -30.0, -15.0, 0.0, 15.0, 30.0, 45.0], theta_list=[60.0, 75.0, 90.0], bounding_box_path='./data/3d_box/doll_glass_bbox.ply', iters=30000, lr=0.0005, sh_degree=0, position_lr_init=0.00016, position_lr_final=1.6e-06, position_lr_delay_mult=0.01, position_lr_max_steps=30000, feature_lr=0.0025, opacity_lr=0.05, scaling_lr=0.005, rotation_lr=0.001, percent_dense=0.01, lambda_dssim=0.2, densification_interval=100, opacity_reset_interval=3000, densify_from_iter=500, densify_until_iter=15000, densify_grad_threshold=0.0002, min_opacity=0.005, weight_decay=0, ckpt='latest', data_path='/mnt/d/dataset/data_DTU/dtu_scan105/', if_data_cuda=True, data_type='dtu', initial_points=None, bg_color=None, R_path='./data/colmap_doll/Orient_R.npy', sample_R_path=None, batch_size=1, batch_rays=512, train_resolution_level=1, eval_resolution_level=4, num_work=0, train_batch_type='image', val_batch_type='image', radius_range=[0.15, 0.15], fovy_range=[50, 70], phi_range=[-180, 180], theta_range=[60, 90]) [INFO] Trainer: df | 2024-04-25_17-27-37 | cuda | fp16 | ./res_gaussion/colmap_doll [INFO] Loading latest checkpoint ... [WARN] No checkpoint found, model randomly initialized. Number of points at initialisation : 1538 0it [00:02, ?it/s] Traceback (most recent call last): File "/home/tom/fsas/xjli/mmlab/TIP-Editor/train_3DGS.py", line 107, in trainer.sample_mask_views(test_loader, os.path.join(opt.workspace, 'sample_views')) File "/home/tom/fsas/xjli/mmlab/TIP-Editor/models/trainer_3dgs.py", line 401, in sample_mask_views preds, preds_depth, dir = self.test_step(data, if_gui=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/tom/fsas/xjli/mmlab/TIP-Editor/models/trainer_3dgs.py", line 280, in test_step outputs = self.model.render(cur_cam, bg_color=self.bg_color) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/tom/fsas/xjli/mmlab/TIP-Editor/models/network_3dgaussain.py", line 815, in render rendered_image, radii, rendered_depth, rendered_alpha = rasterizer( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ValueError: not enough values to unpack (expected 4, got 2)

I did not find the same problem in 3DGS issuses. Could you please give me some help? Thanks.

[zjy526223908]

use this version of gaussian rasterization

install gaussian rasterization

git clone --recursive https://github.com/ashawkey/diff-gaussian-rasterization pip install ./diff-gaussian-rasterization