You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In spdl.io module, there are to_numpy, to_numba and to_torch functions which convert the CPUBuffer and the CUDABuffer into the array format of each framework without copy.
It would be nice if we can extend the functionality to DLPack and Jax.
Requirements:
spdl.io.to_dlpack(buffer: CPUBuffer | CUDABuffer): convert CPUBuffer and CUDABuffer into DLPack native array format without copy-ing.
spdl.io.to_jax(buffer: CPUBuffer | CUDABuffer): convert CPUBuffer and CUDABuffer into Jax native array format without copy-ing.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
In
spdl.io
module, there areto_numpy
,to_numba
andto_torch
functions which convert the CPUBuffer and the CUDABuffer into the array format of each framework without copy.It would be nice if we can extend the functionality to DLPack and Jax.
Requirements:
spdl.io.to_dlpack(buffer: CPUBuffer | CUDABuffer)
: convert CPUBuffer and CUDABuffer into DLPack native array format without copy-ing.spdl.io.to_jax(buffer: CPUBuffer | CUDABuffer)
: convert CPUBuffer and CUDABuffer into Jax native array format without copy-ing.Beta Was this translation helpful? Give feedback.
All reactions