torsh
There is very little structured metadata to build this page
from currently. You should check the
main library docs,
readme, or
Cargo.toml
in case the author documented the features in them.
This version has 61 feature flags, 10 of them enabled by default.
default
data (default)
examples (default)
nn (default)
optim (default)
std (default)
torsh-data (default)
memmap2 (default)
rand (default)
torsh-nn (default)
torsh-optim (default)
backend
bench
cuda
debug
dev
distributed
fast-math
full
full-advanced
functional
fx
hub
jit
linalg
minimal
no_std
numpy
parallel
profiler
python
pytorch-compat
quantization
research
rocm
serialize
serialize-all
serialize-arrow
serialize-hdf5
serialize-onnx
simd
sparse
special
text
torsh-backend
torsh-distributed
torsh-functional
torsh-fx
torsh-hub
torsh-jit
torsh-linalg
torsh-profiler
torsh-quantization
torsh-sparse
torsh-special
torsh-text
torsh-vision
trace
vision
wgpu