train_data samples: '0'

[kg571852741]

train_data samples: '0'

Hi, I have an issue with loading the data when I initiate the model. The train_data samples show '0'. I wonder if my training dataset root config is wrong or if there is another cause for this issue.

SemanticKITTI data

(SphereFormer) bim-group@bimgroup-MS-7D70:~/Documents/GitHub/lightning-hydra-template/SphereFormer$ ls -p data/SemanticKITTI/dataset/sequences/
00/ 01/ 02/ 03/ 04/ 05/ 06/ 07/ 08/ 09/ 10/

Error_logs:

'Traceback (most recent call last): File "train.py", line 908, in <module> main() File "train.py", line 92, in main main_worker(args.train_gpu, args.ngpus_per_node, args) File "train.py", line 294, in main_worker collate_fn=collate_fn File "/home/bim-group/anaconda3/envs/SphereFormer/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 266, in __init__ sampler = RandomSampler(dataset, generator=generator) # type: ignore File "/home/bim-group/anaconda3/envs/SphereFormer/lib/python3.7/site-packages/torch/utils/data/sampler.py", line 104, in __init__ "value, but got num_samples={}".format(self.num_samples)) ValueError: num_samples should be a positive integer value, but got num_samples=0'

Log.txt

[08/0907:31:59] main-logger INFO: a: 0.0125 arch: unet_spherical_transformer aug: True base_lr: 0.006 batch_size: 4 batch_size_val: 4 block_reps: 2 block_residual: True class_weight: [3.1557, 8.7029, 7.8281, 6.1354, 6.3161, 7.9937, 8.9704, 10.1922, 1.6155, 4.2187, 1.9385, 5.5455, 2.0198, 2.6261, 1.3212, 5.1102, 2.5492, 5.8585, 7.3929] classes: 19 data_name: semantic_kitti data_root: data/SemanticKITTI/dataset dist_backend: nccl dist_url: tcp://127.0.0.1:6789 distributed: False drop_path_rate: 0.3 drop_rate: 0.5 epochs: 50 eval_freq: 1 evaluate: True fea_dim: 6 grad_checkpoint_layers: [] ignore_label: 255 input_c: 4 label_mapping: util/semantic-kitti.yaml layers: [32, 64, 128, 256, 256] loss_name: ce_loss m: 32 manual_seed: 123 max_batch_points: 1000000 momentum: 0.9 multiprocessing_distributed: False ngpus_per_node: 2 patch_size: [0.05 0.05 0.05] pc_range: [[-51.2, -51.2, -4], [51.2, 51.2, 2.4]] power: 0.9 print_freq: 10 quant_size_scale: 24 rank: 0 rel_key: True rel_query: True rel_value: True resume: None save_freq: 1 save_path: runs/semantic_kitti_unet32_spherical_transformer scheduler: Poly scheduler_update: step sphere_layers: [1, 2, 3, 4, 5] start_epoch: 0 sync_bn: True train_gpu: [0, 1] transformer_lr_scale: 0.1 use_amp: True use_tta: False use_xyz: True val: False vote_num: 4 voxel_max: 120000 voxel_size: [0.05, 0.05, 0.05] weight: None weight_decay: 0.02 window_size: 6 window_size_scale: [2.0, 1.5] window_size_sphere: [2, 2, 80] workers: 32 world_size: 1 xyz_norm: False [08/09 07:32:00] main-logger INFO: => creating model ... [08/09 07:32:00] main-logger INFO: Classes: 19 ... [08/09 07:32:01] main-logger INFO: #Model parameters: 32311715 [08/09 07:32:01] main-logger INFO: class_weight: tensor([ 3.1557, 8.7029, 7.8281, 6.1354, 6.3161, 7.9937, 8.9704, 10.1922, 1.6155, 4.2187, 1.9385, 5.5455, 2.0198, 2.6261, 1.3212, 5.1102, 2.5492, 5.8585, 7.3929], device='cuda:0') [08/09 07:32:01] main-logger INFO: loss_name: ce_loss [08/09 07:32:01] main-logger INFO: train_data samples: '0'

[X-Lai]

I suspect whether a wrong data path is passed. Can you double check it?

[ccdont]

Hi, I had the same problem, have you solved it yet?