Skip to content

Avoid using numba to set device context in import #145

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
karlhigley merged 6 commits into from
Oct 5, 2022
Merged

Avoid using numba to set device context in import #145

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
67f2df1
avoid numba context in import
rjzamora Oct 4, 2022
d6fb5a9
use pynvml instead of dask-cuda
rjzamora Oct 4, 2022
18c42f3
reduce diff
rjzamora Oct 4, 2022
e9e750b
pylint
rjzamora Oct 4, 2022
9efdbf9
use distributed.nvml
rjzamora Oct 4, 2022
6843832
Merge branch 'main' into has_gpu-fix
karlhigley Oct 5, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 11 additions & 5 deletions merlin/core/compat.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,13 +13,19 @@
# See the License for the specific language governing permissions and
# limitations under the License.
#
HAS_GPU = False
try:
from numba import cuda

try:
HAS_GPU = len(cuda.gpus.lst) > 0
except cuda.cudadrv.error.CudaSupportError:
pass
except ImportError:
cuda = None

HAS_GPU = False
try:
from dask.distributed.diagnostics import nvml

HAS_GPU = nvml.device_get_count() > 0
except ImportError:
# We can use `cuda` to set `HAS_GPU` now that we
# know `distributed` is not installed (otherwise
# the `nvml` import would have succeeded)
HAS_GPU = cuda is not None