[cuTENSOR] Automatically enable/disable TensorFloat32 #40
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Linux-cpp-lisp
changed the title
gsakhnovsky-nvidia
force-pushed
the
master
branch
from
ec627be to
5c03ab7
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
(Thanks to @springer13 for making your work on a PyTorch cuTENSOR wrapper public!)
Currently, the Python cuTENSOR wrapper always uses TensorFloat32 as the compute dtype for 32-bit float tensors, which is unsupported on non-Ampere GPUs. This PR uses the PyTorch and Tensorflow configuration options for TensorFloat32 (which autodetect Ampere) to set the compute dtype to normal 32-bit float when tf32 is not supported.