Hi All!
I used:
pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/xpu
into a freshly-created python=3.13 conda environment. No installation errors were reported
and the pytorch version appears to be as expected, but xpu support appears to be absent:
(2_8_0_xpu_nightly) xxxxx@xxxxxxxxxxx:~/Documents/devl$ python
Python 3.13.2 | packaged by Anaconda, Inc. | (main, Feb 6 2025, 18:56:02) [GCC 11.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.__version__
'2.8.0.dev20250325+xpu'
>>> tt = torch.randn (10, device = 'xpu')
Traceback (most recent call last):
File "<python-input-2>", line 1, in <module>
tt = torch.randn (10, device = 'xpu')
File "/home/xxxxx/miniconda3/envs/2_8_0_xpu_nightly/lib/python3.13/site-packages/torch/xpu/__init__.py", line 118, in _lazy_init
raise AssertionError("Torch not compiled with XPU enabled")
AssertionError: Torch not compiled with XPU enabled
(I get the same result in a python=3.12 environment, while the release version, ‘2.6.0+xpu’
works,)
Is there any way I can fix this, or should I just wait until, say, a 2.7 xpu-enabled version is
released?
Thanks.
K. Frank