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
{{ message }}
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
before the PR #17123split_and_load when passed CSRNDArray would return CSRNDArray for batch_axis=0 but now return NDArray.
To Reproduce
import scipy.sparse as sps
import mxnet.ndarray.sparse as mxsps
import mxnet as mx
x = mxsps.csr_matrix(sps.coo_matrix(([2.0], ([99], [999]))).tocsr(), ctx=mx.gpu(0))
y = mx.gluon.utils.split_and_load(x, (mx.gpu(0), mx.gpu(0)))
print(x)
print(y)
Output:
ubuntu@ip-172-31-90-243 ~/workspace/incubator-mxnet (mx1x) $ python debug.py
[20:54:45] ../src/base.cc:84: Upgrade advisory: this mxnet has been built against cuDNN lib version 7501, which is older than the oldest version tested by CI (7600). Set MXNET_CUDNN_LIB_CHECKING=0 to quiet this warning.
<CSRNDArray 100x1000 @gpu(0)>
[
<NDArray 50x1000 @gpu(0)>,
<NDArray 50x1000 @gpu(0)>]
Description
before the PR #17123
split_and_load
when passed CSRNDArray would return CSRNDArray for batch_axis=0 but now return NDArray.To Reproduce
Output:
What have you tried to solve it?
PR for Fix #19267
Environment
We recommend using our script for collecting the diagnositc information. Run the following command and paste the outputs below:
The text was updated successfully, but these errors were encountered: