목록pyTorch (6)

Shakerato

pytorch CUDNN_STATUS_EXECUTION_FAILED error (Anaconda3)

[previous setup]conda install pytorch cuda90 -c pytorch (installed)pytorch 0.4.0 py36_cuda90_cudnn7he774522_1 [cuda90] pytorch --> If you are using the latest GPU in the above environment, for example, 2080ti,you may get an error message when you run pytorch code using GPU as shown in below. CUDNN_STATUS_EXECUTION_FAILED --> [solution setup]1. install cuda 10.0 and new cudnn for this2. conda upd..
Research 2019. 2. 23. 20:37
Video Action Classification using 3D ResNet

Reference:https://github.com/kenshohara/video-classification-3d-cnn-pytorch (Linux)1. Install Anaconda and Setup Pytorch (GPU)2. Download code from Github (https://github.com/kenshohara/video-classification-3d-cnn-pytorch)3. Download Pre-trained Model and copy to source code folder (resnext-101-kinetics.pth) https://drive.google.com/drive/folders/1zvl89AgFAApbH0At-gMuZSeQB_LpNP-M4. Install FFmpe..
Research 2019. 2. 18. 12:13
GPU devices checking tips for Deep Learning (Be careful!)

[Tensorflow]from tensorflow.python.client import device_libprint(device_lib.list_local_devices()) [Keras]from keras import backend as KK.tensorflow_backend._get_available_gpus() [Pytorch]import torchtorch.cuda.get_device_name(0) # number of gpu (cuda:0)torch.cuda.get_device_name(1) # number of gpu (cuda:1) Be careful with the above codes! The code for checking device(GPU) takes your GPU memory a..
Research 2019. 1. 16. 17:44
pytorch "RuntimeError: ~ inplace operation" for CUDA

def __init__(self, input_size, hidden_size, num_layers=2): super().__init__() self.lstm = nn.LSTM(input_size, hidden_size, num_layers) def forward(self, features, init_hidden=None): self.lstm.flatten_parameters() RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation Move the self.lstm.flatten_parameters() in def foward() to def __init__. It ..
Research 2019. 1. 5. 17:58
Install PyTorch using Anaconda on Windows 10

1. check cuda9.0 installed (nvcc --version) (if yet, install cuda) https://developer.nvidia.com/cuda-90-download-archive (if no gpu) pass this step 2, Install Anaconda https://www.anaconda.com/download/ 3. in 'cmd' conda create -n pytorch python=3.6 anaconda (follow https://anaconda.org/soumith/pytorch)4. in 'cmd' conda install pytorch cuda90 -c pytorch (if no gpu) conda install pytorch-cpu -c p..
Research 2018. 6. 10. 19:50
Google colab에서 한국어 NLP 처리 환경 구축

What is Colab (Colaboratory) ?- https://colab.research.google.com- Colaboratory는 기계 학습 교육 및 연구 보급을 돕기 위해 만든 Google 연구 프로젝트입니다.- Jupyter 노트북 환경에서 클라우드 환경에서의 설치 및 실행이 전혀 필요하지 않습니다. - Colaboratory 노트는 Google 드라이브에 저장되며 Google 문서 또는 스프레드 시트에서와 마찬가지로 공유 할 수 있습니다.- Colaboratory는 무료입니다.- Now you can develop deep learning applications with Google Colaboratory on the free Tesla K80 GPU- using Keras, Tens..
Research 2018. 3. 31. 20:35
Prev 1 Next