gpu memory is increasing as training

[blue-butterfly]

Hi, thanks for your great work ! But when I train as the README, the gpu memory is increasing， Then out of memory ？ The commands as follow：

cd cutler export DETECTRON2_DATASETS=/path/to/DETECTRON2_DATASETS/ python train_net.py --num-gpus 2 \ --config-file model_zoo/configs/CutLER-ImageNet/cascade_mask_rcnn_R_50_FPN.yaml

[frank-xwang]

Hi! I recommend using 8 GPUs to run the experiments. However, suppose you are limited to 2 GPUs. In that case, you may need to adjust the batch size, learning rate, and the number of iterations accordingly to avoid encountering an out-of-memory error.

I am closing this issue now. Please feel free to reopen it if you have more questions.

[khatanas]

Hi!

• Where can we change these parameters from ?
• How many GiB of GPU were available while training on 8GPUs ?

Thank you in advance

[frank-xwang]

You can change "--num-gpus" in the training command. You can modify other parameters in the "--config-file" you specify.

[khatanas]

Perfect! Thank you 👍