When I recreate and reassign a JAX np array to the same variable name, for some reason the GPU memory nearly doubles the first recreation and then stays stable for subsequent recreations/reassignments. Why does this happen and is this generally expected behavior for JAX arrays? Fully runnable minimal example:…
Tag: gpu
pytorch cuda out of memory while inferencing
I think this is a very basic question, my apologies as I am very new to pytorch. I am trying to find if an image is manipulated or not using MantraNet. After running 2-3 inferences I get the CUDA out of memory, then after restarting the kernel also I keep getting the same error: The error is given below: Runt…
Why tensor size was not changed?
I made the toy CNN model. Then, I had checked model.summary via this code And I was able to get the following results: I want to reduce model size cuz i wanna increase the batch size. So, I had changed torch.float32 -> torch.float16 via NVIDIA/apex As a result, torch.dtype was changed torch.float16 from to…
Cannot install the gpu version of torch and torchvision in poetry due to a dependency problem
I am trying to create a virtual environment for machine learning using poetry. So, I am using pytorch as a framework for deep learning. I will extract the relevant part of my pyproject.toml. Since pytroch uses the GPU, you need to install it by specifying the whl file. If you install it this way, the version …
TypeError: expected CPU (got CUDA)
X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.2,random_state=0) y_train import torch import torch.nn as nn import torch.nn.functional as F when I Run this code I got this error: How to can I solve this error? Answer To transfer the variables to GPU, try the following:
Floyd-Warshall algorithm on GPU using numba
I’m writing optimised Floyd-Warshall algorithm on GPU using numba. I need it to work in a few seconds in case of 10k matricies. Right now the processing is done in around 60s. Here is my code: To be honest I’m pretty new to writing scripts on GPU, so do you have any ideas how to make this code eve…
ERROR: Could not find a version that satisfies the requirement dask-cudf (from versions: none)
Describe the bug When I am trying to import dask_cudf I get the following ERROR: I have dask and RAPIDS installed with pip when I search for: pip install dask_cudf original site is not exists anymore: https://pypi.org/project/dask-cudf/ google stored site history: https://webcache.googleusercontent.com/search…
tensorflow cannot find GPU
I had install “tensorflow-GPU”, CUDA 10.0. and my GPU is GTX1660 ti. I also tested bu CUDA 10.2 and 11. I added cudnn to windows PATH but I still got this error. Answer I found the problem. the problem was versions of CUDA and cudnn.
Get LightGBM/ LGBM run with GPU on Google Colabratory
I often run LGBM on Google Colabratory and I just found out this page saying that LGBM it set to CPU by default so you need to set up first. https://medium.com/@am.sharma/lgbm-on-colab-with-gpu-c1c09e83f2af So I executed the code recommended on the page or some other codes recommended on stackoverflow as foll…
PyTorch embedding layer raises “expected…cuda…but got…cpu” error
I’m working on translating a PyTorch model from CPU (where it works) to GPU (where it so far doesn’t). The error message (clipped to the important bits) is as follows: Here is the full model definition: This type of error typically occurs when there is a tensor in the model that should be on GPU b…