Some error during training, please help about pgcn HOT 3 CLOSED

Could you provide more details about the error?

from pgcn.

(learnpytorch) C:\Users\gaura\OneDrive\Documents\GitHub\pgcn>python pgcn_train.py thumos14 --snapshot_pre /Users/gaura/OneDrive/Desktop/BHU_Intern
2020-05-15 19:33:15,384 INFO
creating folder: /Users/gaura/OneDrive/Desktop/BHU_Intern
|===>Backups using time: 2.321 s
|===>Backups will be saved at /Users/gaura/OneDrive/Desktop/BHU_Intern
2020-05-15 19:33:17,723 INFO
runtime args

Namespace(batch_size=32, clip_gradient=None, comp_loss_weight=0.5, dataset='thumos14', dataset_configs=ordereddict([('train_ft_path', '/Users/gaura/OneDrive/Desktop/BHU_Intern/Flow_Train_All'), ('test_ft_path', '/Users/gaura/OneDrive/Desktop/BHU_Intern/Flow_Test_All'), ('train_dict_path', 'data/thumos14_train_prop_dict.pkl'), ('val_dict_path', 'data/thumos14_val_prop_dict.pkl'), ('test_dict_path', 'data/thumos14_test_prop_dict.pkl'), ('train_prop_file', 'data/bsn_train_proposal_list.txt'), ('test_prop_file', 'data/bsn_test_proposal_list.txt'), ('training_epoch_multiplier', 20), ('testing_epoch_multiplier', 1), ('fg_iou_thresh', 0.7), ('bg_iou_thresh', 0.01), ('incomplete_iou_thresh', 0.3), ('bg_coverage_thresh', 0.02), ('incomplete_overlap_thresh', 0.01), ('prop_per_video', 8), ('fg_ratio', 1), ('bg_ratio', 1), ('incomplete_ratio', 6), ('iou_threshold', 0.7), ('dis_threshold', 0), ('starting_ratio', 0.5), ('ending_ratio', 0.5)]), dropout=0.8, epochs=70, eval_freq=5, evaluate=False, evaluation=ordereddict([('top_k', 2000), ('nms_threshold', 0.2)]), gpus=None, graph_configs=ordereddict([('adj_num', 21), ('child_num', 4), ('iou_num', 8), ('dis_num', 2)]), iter_size=1, lr=0.01, lr_steps=[15, 30, 45], model_configs=ordereddict([('num_class', 20), ('act_feat_dim', 1024), ('comp_feat_dim', 3072), ('dropout', 0.8), ('gcn_dropout', 0.7)]), momentum=0.9, print_freq=20, reg_loss_weight=0.5, resume='', snapshot_pref='/Users/gaura/OneDrive/Desktop/BHU_Intern', start_epoch=0, training_epoch_multiplier=10, weight_decay=0.0005, workers=4)

config

ordereddict([('train_ft_path', '/Users/gaura/OneDrive/Desktop/BHU_Intern/Flow_Train_All'), ('test_ft_path', '/Users/gaura/OneDrive/Desktop/BHU_Intern/Flow_Test_All'), ('train_dict_path', 'data/thumos14_train_prop_dict.pkl'), ('val_dict_path', 'data/thumos14_val_prop_dict.pkl'), ('test_dict_path', 'data/thumos14_test_prop_dict.pkl'), ('train_prop_file', 'data/bsn_train_proposal_list.txt'), ('test_prop_file', 'data/bsn_test_proposal_list.txt'), ('training_epoch_multiplier', 20), ('testing_epoch_multiplier', 1), ('fg_iou_thresh', 0.7), ('bg_iou_thresh', 0.01), ('incomplete_iou_thresh', 0.3), ('bg_coverage_thresh', 0.02), ('incomplete_overlap_thresh', 0.01), ('prop_per_video', 8), ('fg_ratio', 1), ('bg_ratio', 1), ('incomplete_ratio', 6), ('iou_threshold', 0.7), ('dis_threshold', 0), ('starting_ratio', 0.5), ('ending_ratio', 0.5)])
File parsed. Time:12.89
100%|################################################################################| 199/199 [08:36<00:00, 2.60s/it]
Dict constructed. Time:519.94
Traceback (most recent call last):
File "pgcn_train.py", line 369, in
main()
File "pgcn_train.py", line 84, in main
test_mode=False),
File "C:\Users\gaura\OneDrive\Documents\GitHub\pgcn\pgcn_dataset.py", line 201, in init
pickle.dump([self.act_iou_dict, self.act_dis_dict, self.prop_dict], open(self.prop_dict_path, "wb"))
MemoryError

from pgcn.

You can monitor the usage of CPU memory. This error seems to be related to the usage of memory.
Make sure that you have enough memory to save the dictionary.

from pgcn.