AlphaTracker is a computer vision pipeline with the practical and real-time advantages , which requires minimal hardware requirements and produces reliable tracking of multiple unmarked animals. An easy-to-use user interface further enables manual inspection and curation of results.
61
stars
16
forks
source link
Error in training: cannot load `AlphaTracker/Tracking/AlphaTracker/models/sppe/duc_se.pth` #18
I obtain the following error when running train.py. Some googling suggests this results from a corrupt file download. When I run md5sum on duc_se.pth, I get 3764b858a2f612f536c841c971ea8e63 (and the file appears to be 34K). Is that correct? If not, can you provide an alternative way to download it?
*** training sppe ***
training with following setting:
CUDA_VISIBLE_DEVICES=0 python train.py \
--dataset coco \
--img_folder_train /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/train_yolo/darknet//data/top/color/ \
--annot_file_train /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/train_sppe//data/top/data_newLabeled_01_train.h5 \
--img_folder_val /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/train_yolo/darknet//data/top/color/ \
--annot_file_val /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/train_sppe//data/top/data_newLabeled_01_val.h5 \
--expID top \
--nClasses 7 \
--LR 0.0001 --trainBatch 10 \
--nEpochs 10 \
--nThreads 6 \
--loadModel /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/models/sppe/duc_se.pth
Loading Model from /n/groups/datta/caleb/22_2_1_keypointSORT/log/22_8_5_alphaTracker/AlphaTracker/Tracking/AlphaTracker/models/sppe/duc_se.pth
Traceback (most recent call last):
File "train.py", line 204, in <module>
main()
File "train.py", line 122, in main
weight_save = torch.load(opt.loadModel)
File "/home/csw14/miniconda3/envs/alphatracker/lib/python3.8/site-packages/torch/serialization.py", line 608, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "/home/csw14/miniconda3/envs/alphatracker/lib/python3.8/site-packages/torch/serialization.py", line 777, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
_pickle.UnpicklingError: invalid load key, '<'.
I obtain the following error when running
train.py
. Some googling suggests this results from a corrupt file download. When I runmd5sum
onduc_se.pth
, I get3764b858a2f612f536c841c971ea8e63
(and the file appears to be 34K). Is that correct? If not, can you provide an alternative way to download it?