RuntimeError: cannot reshape tensor of 0 elements into shape [1, 0, -1] because the unspecified dimension size -1 can be any value and is ambiguous

[Tuo-Liang]

I tried to run the script "bash dev_scripts/ete/dtu_dgt_d012_img0123_conf_color_dir_agg2.sh", but it gave back the following results. I just wonder why this occurred and how to solve. Please help me.

/home/lab710/anaconda3/envs/pointnerf/lib/python3.8/site-packages/numpy/core/shape_base.py:420: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.
arrays = [asanyarray(arr) for arr in arrays]
dataset total: train 30184
dataset [DtuDataset] was created

training images = 30184

No previous checkpoints, start from scratch!!!!
opt.resume_dir ../checkpoints/init/init/dtu_dgt_d012_img0123_conf_color_dir_agg2 None
opt.act_type!!!!!!!!! LeakyReLU
querier device cuda:0 0
no neural points as nn.Parameter
model [MvsPointsVolumetricModel] was created
opt.resume_iter!!!!!!!!! None
loading mvs from ../checkpoints/init/init/dtu_dgt_d012_img0123_conf_color_dir_agg2/None_net_mvs.pth
cannot load ../checkpoints/init/init/dtu_dgt_d012_img0123_conf_color_dir_agg2/None_net_mvs.pth
loading ray_marching from ../checkpoints/init/init/dtu_dgt_d012_img0123_conf_color_dir_agg2/None_net_ray_marching.pth
cannot load ../checkpoints/init/init/dtu_dgt_d012_img0123_conf_color_dir_agg2/None_net_ray_marching.pth
------------------- Networks -------------------
[Network mvs] Total number of parameters: 0.041M
[Network ray_marching] Total number of parameters: 0.397M

/home/lab710/anaconda3/envs/pointnerf/lib/python3.8/site-packages/torch/functional.py:507: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3549.)
return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
points_dirs torch.Size([1, 0, 3])
1 0

Traceback (most recent call last):
File "train.py", line 366, in
main()
File "train.py", line 264, in main
model.optimize_parameters(total_steps=total_steps)
File "/home/lab710/pointnerf/run/../models/neural_points_volumetric_model.py", line 215, in optimize_parameters
self.forward()
File "/home/lab710/pointnerf/run/../models/mvs_points_volumetric_model.py", line 123, in forward
points_xyz, points_embedding, points_colors, points_dirs, points_conf = self.net_mvs(self.input)
File "/home/lab710/anaconda3/envs/pointnerf/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/home/lab710/anaconda3/envs/pointnerf/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "/home/lab710/pointnerf/run/../models/mvs/mvs_points_model.py", line 372, in forward
points_features_lst = [self.query_embedding(HDWD, torch.as_tensor(cam_xyz_lst[i][None, ...], device="cuda", dtype=torch.float32), photometric_confidence_lst[i][None, ..., None], img_feats, data_mvs['c2ws'], data_mvs['w2cs'], batch["intrinsics"], int(self.args.depth_vid[i]), pointdir_w=False) for i in range(len(cam_xyz_lst))]
File "/home/lab710/pointnerf/run/../models/mvs/mvs_points_model.py", line 372, in
points_features_lst = [self.query_embedding(HDWD, torch.as_tensor(cam_xyz_lst[i][None, ...], device="cuda", dtype=torch.float32), photometric_confidence_lst[i][None, ..., None], img_feats, data_mvs['c2ws'], data_mvs['w2cs'], batch["intrinsics"], int(self.args.depth_vid[i]), pointdir_w=False) for i in range(len(cam_xyz_lst))]
File "/home/lab710/pointnerf/run/../models/mvs/mvs_points_model.py", line 253, in query_embedding
points_dirs = points_dirs.view(cam_xyz.shape[0], cam_xyz.shape[1], -1)
RuntimeError: cannot reshape tensor of 0 elements into shape [1, 0, -1] because the unspecified dimension size -1 can be any value and is ambiguous
end loading

PyCUDA ERROR: The context stack was not empty upon module cleanup.

A context was still active when the context stack was being
cleaned up. At this point in our execution, CUDA may already
have been deinitialized, so there is no way we can finish
cleanly. The program will be aborted now.
Use Context.pop() to avoid this problem.

dev_scripts/ete/dtu_dgt_d012_img0123_conf_color_dir_agg2.sh: line 223: 41130 Aborted (core dumped) python3 train.py --experiment $name --data_root $data_root --dataset_name $dataset_name --model $model --which_render_func $which_render_func --which_blend_func $which_blend_func --out_channels $out_channels --num_pos_freqs $num_pos_freqs --num_viewdir_freqs $num_viewdir_freqs --random_sample $random_sample --random_sample_size $random_sample_size --batch_size $batch_size --maximum_step $maximum_step --lr $lr --lr_policy $lr_policy --lr_decay_iters $lr_decay_iters --gpu_ids $gpu_ids --checkpoints_dir $checkpoints_dir --save_iter_freq $save_iter_freq --niter $niter --niter_decay $niter_decay --n_threads $n_threads --pin_data_in_memory $pin_data_in_memory --train_and_test $train_and_test --test_num $test_num --test_freq $test_freq --test_num_step $test_num_step --test_color_loss_items $test_color_loss_items --print_freq $print_freq --bg_color $bg_color --split $split --which_ray_generation $which_ray_generation --near_plane $near_plane --far_plane $far_plane --dir_norm $dir_norm --which_tonemap_func $which_tonemap_func --load_points $load_points --resume_dir $resume_dir --resume_iter $resume_iter --feature_init_method $feature_init_method --agg_axis_weight $agg_axis_weight --agg_distance_kernel $agg_distance_kernel --radius_limit_scale $radius_limit_scale --depth_limit_scale $depth_limit_scale --vscale $vscale --kernel_size $kernel_size --SR $SR --K $K --P $P --NN $NN --agg_feat_xyz_mode $agg_feat_xyz_mode --agg_alpha_xyz_mode $agg_alpha_xyz_mode --agg_color_xyz_mode $agg_color_xyz_mode --save_point_freq $save_point_freq --raydist_mode_unit $raydist_mode_unit --agg_dist_pers $agg_dist_pers --agg_intrp_order $agg_intrp_order --shading_feature_mlp_layer1 $shading_feature_mlp_layer1 --shading_feature_mlp_layer2 $shading_feature_mlp_layer2 --shading_feature_mlp_layer3 $shading_feature_mlp_layer3 --shading_feature_num $shading_feature_num --dist_xyz_freq $dist_xyz_freq --shpnt_jitter $shpnt_jitter --shading_alpha_mlp_layer $shading_alpha_mlp_layer --shading_color_mlp_layer $shading_color_mlp_layer --which_agg_model $which_agg_model --num_feat_freqs $num_feat_freqs --dist_xyz_deno $dist_xyz_deno --apply_pnt_mask $apply_pnt_mask --point_features_dim $point_features_dim --color_loss_items $color_loss_items --color_loss_weights $color_loss_weights --feedforward $feedforward --trgt_id $trgt_id --depth_vid $depth_vid --ref_vid $ref_vid --manual_depth_view $manual_depth_view --pre_d_est $pre_d_est --depth_occ $depth_occ --manual_std_depth $manual_std_depth --visual_items $visual_items --appr_feature_str0 $appr_feature_str0 --appr_feature_str1 $appr_feature_str1 --appr_feature_str2 $appr_feature_str2 --appr_feature_str3 $appr_feature_str3 --act_type $act_type --point_conf_mode $point_conf_mode --point_dir_mode $point_dir_mode --point_color_mode $point_color_mode --depth_conf_thresh $depth_conf_thresh --geo_cnsst_num $geo_cnsst_num --bgmodel $bgmodel --vox_res $vox_res --debug

[iszhihao]

I have the same issue

[Tuo-Liang]

I have the same issue

do you know how to solve it ? I found that it might cause by the "cam_xyz" got 'nan". But I dont know how to solve it.

[iszhihao]

I have the same issue

do you know how to solve it ? I found that it might cause by the "cam_xyz" got 'nan". But I dont know how to solve it.

The problem has not been resolved yet. When I run other .sh files, such as chair.sh and lego.sh, I also encounter a "segmentation fault" error, which is driving me to the brink of frustration.

[iszhihao]

I have the same issue

do you know how to solve it ? I found that it might cause by the "cam_xyz" got 'nan". But I dont know how to solve it.

do you know how to solve it