Getting the below error when trying to run my experiment (loading dataset) in GPU. My training and development data set sizes are ( 252000 train.txt, and 28000 devel.txt). Now working on a single GPU.
X_temp = X_temp.cuda()
RuntimeError: CUDA error: out of memory
What is the X_temp.size() ? how much memory do you have in your GPU ? did you check if another program is using the GPU and how much memory do you have left (i.e. nvidia-smi in the terminal) ?