PyTorch
PyTorch is an open-source machine learning framework that provides tensor computation with GPU acceleration and automatic differentiation for building and training deep neural networks.
Available Pages
PyTorch - The Deep Learning Framework That Doesn't Suck
PyTorch: Intuitive API, effective debugging. Learn why it's preferred over TensorFlow. Get solutions for common GPU memory and `torch.compile` problems.
PyTorch Production Deployment: Scale, Optimize & Prevent Crashes
Learn the brutal truth about PyTorch production deployment. Move models from notebooks to scale, optimize performance, prevent OOM crashes, and monitor real-world traffic.
PyTorch Debugging & Troubleshooting Guide: Fix Common Model Errors
Master PyTorch debugging with essential tools and advanced techniques. Learn to resolve cryptic errors like 'RuntimeError' and 'CUDA assert triggered' for robust model development.
Related Technologies
Competition
Integration
cuda
Official integration support
rocm
Official integration support
tensorrt
Official integration support
onnx
Official integration support
hugging face
Official integration support
mlflow
Official integration support
weights and biases
Official integration support
ray
Official integration support
docker
Official integration support
kubernetes
Official integration support
aws sagemaker
Official integration support
azure ml
Official integration support
google cloud ai
Official integration support
Dependencies
python
Foundation technology
numpy
Requires for operation
cpp
Foundation technology
openmp
Requires for operation
intel mkl
Requires for operation
cudnn
Requires for operation
nccl
Requires for operation
Development
torchvision
Adds functionality to
torchaudio
Adds functionality to
torchtext
Adds functionality to
torchserve
Adds functionality to
pytorch lightning
Adds functionality to
captum
Adds functionality to
pytorch geometric
Adds functionality to