mirror of
https://github.com/NohamR/Stage-2024.git
synced 2026-01-11 16:58:23 +00:00
test setup
This commit is contained in:
63
yolov7-setup/info_SF.txt
Normal file
63
yolov7-setup/info_SF.txt
Normal file
@@ -0,0 +1,63 @@
|
||||
Dépôt :
|
||||
git clone https://github.com/JackWoo0831/Yolov7-tracker
|
||||
|
||||
mv Yolov7-tracker yolov7-tracker
|
||||
cd yolov7-tracker
|
||||
git checkout v2 # change to v2 branch !!
|
||||
|
||||
conda create -n yolov7 python=3.9 pytorch=1.12 ## ancienne version, comme indiquée dans le readme, sinon numpy.distutils manque...
|
||||
conda activate yolov7
|
||||
|
||||
( si besoin : conda remove --name yolov7 --all )
|
||||
|
||||
(yolov7) 14:37:21 sylvain.faure@cinaps ~/RN/athle/yolov7-tracker $ pip3 install numpy scipy matplotlib cython pandas cuda-python
|
||||
(yolov7) 15:02:25 sylvain.faure@cinaps ~/RN/athle/yolov7-tracker $ pip3 install -r requirements.txt
|
||||
|
||||
(yolov7) 15:02:25 sylvain.faure@cinaps ~/RN/athle/yolov7-tracker $ pip3 install ultralytics==8.0.94
|
||||
|
||||
#################
|
||||
## TRAIN
|
||||
#################
|
||||
|
||||
## entrainement avec un premier dataset : ~/RN/athle/yolov7-tracker/dataset1_2024_06_19/
|
||||
Deux fichiers de configuration créées :
|
||||
tracker/config_files/dataset1_2024_06_19.yaml
|
||||
data/files_dataset1_2024_06_19.yaml
|
||||
|
||||
yolov7 sur node20 du cluster :
|
||||
|
||||
## pour voir la version de CUDA du noeud :
|
||||
(yolov7) sylvain.faure@node20:~$ nvidia-smi
|
||||
=> cuda 11.4 sur node20
|
||||
|
||||
cf notes d'installation de pytorch pour trouver les bonnes versions : https://pytorch.org/get-started/previous-versions/
|
||||
pip3 install torch==2.2.2 torchvision==0.17.2 torchaudio==2.2.2 --index-url https://download.pytorch.org/whl/cu118
|
||||
pip3 install filterpy
|
||||
|
||||
## test pour verifier que cuda est bien ok :
|
||||
>>> import torch
|
||||
>>> torch.cuda.is_available()
|
||||
True
|
||||
|
||||
## lancement de l'entrainement :
|
||||
|
||||
(yolov7) sylvain.faure@node20:~/RN/athle/yolov7-tracker$ python3 train.py --dataset dataset1_2024_06_19__ --workers 1 --device 0 --batch-size 4 --d
|
||||
ata data/dataset1_2024_06_19/dataset.yaml --img 1280 720 --cfg cfg/training/yolov7x_dataset1_2024_06_19.yaml --weights '' --name yolov7x-dataset1_
|
||||
2024_06_19 --hyp data/hyp.scratch.custom.yaml
|
||||
|
||||
|
||||
Si erreur : _pickle.UnpicklingError: STACK_GLOBAL requires str
|
||||
Effacer les fichiers .cache : liste_images.cache par exemple...
|
||||
(yolov7) sylvain.faure@node20:~/RN/athle/yolov7-tracker$rm data/dataset1_2024_06_19/liste_images.cache
|
||||
|
||||
Si erreur np.int => remplacer par des np.int64
|
||||
|
||||
Si erreur cuda/cpu dans loss.py :
|
||||
you have to replace the line in the file yolo7/utils/loss.py
|
||||
"from_which_layer.append((torch.ones(size=(len(b),)) * i)"
|
||||
to "from_which_layer.append((torch.ones(size=(len(b),)) * i).to('cuda'))",
|
||||
and add new line "fg_mask_inboxes = fg_mask_inboxes.to(torch.device('cuda'))"
|
||||
after "fg_mask_inboxes = matching_matrix.sum(0) > 0.0"
|
||||
so you need to do it 3 times in the file
|
||||
|
||||
ssh cinaps
|
||||
Reference in New Issue
Block a user