collate_fn for dataloader and extract vision features
484e90b
tcm03
commited on
Commit History
Segment long videos and multithreading in EnTubeDataset
51273ab
tcm03
commited on
Analyze running time of process_images() and process_video_frames()
fc48d96
tcm03
commited on
Remove HF home
cc6eb9f
tcm03
commited on
inspect VideoReader time
18edb19
tcm03
commited on
Debugging learn process_images
040ba84
tcm03
commited on
Restructure repo and refine annotation
597f824
tcm03
commited on
Fix pickle file name
5a1acba
tcm03
commited on
Add memory snapshot
6e715c9
tcm03
commited on
Debug EnTubeDataset
59dc16e
tcm03
commited on
I think batching data loader is more general: give batch_size=1 in case limited mem
b231d57
tcm03
commited on
Back to multithreading for individual sample
1098248
tcm03
commited on
Debug dataloader
ddef400
tcm03
commited on
debug process_images
f834242
tcm03
commited on
debug process_images
d973ae3
tcm03
commited on
debug process_images
075ba26
tcm03
commited on
break
29d12fa
tcm03
commited on
Debug process_images
9473b22
tcm03
commited on
Debug process_images
57e99b9
tcm03
commited on
Fix impor
9aabd7c
tcm03
commited on
Rename dataset.py to entube_dataset.py
7d59f30
tcm03
commited on
Update code
cd45874
tcm03
commited on
remove self arg
a0c9d15
tcm03
commited on
add import json in preprocessor.py
2b91970
tcm03
commited on
Add config_file arg
5696d8d
tcm03
commited on
Add import torch in preprocessor.py
bd72620
tcm03
commited on
Add preprocessing module to pre-extract vision features
9d38064
tcm03
commited on