-
Auto detect KVS port for oneCCL to avoid port conflict
xwu99 updated
3 years ago
-
Use getifaddrs instead of "hostname -I"
xwu99 updated
3 years ago
-
In our automatic tests for OAP product, we found that using Intel-MLlib to run PCA and Kmeans algorithms often meet hanging issue which lead to block the whole workflow. The phenomenon is shown in the…
-
xwu99 updated
3 years ago
-
Hey,
I'm using oneCCL to implement multinode communication in the marian machine translation toolkit. I am having a problem with a call to `ccl::allgatherv`. As far as I can tell according to the d…
-
error message:
```
/tmp/pip-req-build-c3bi1v9x/third_party/torch_ccl/src/./init.h:34:10: fatal error: pybind11/pybind11.h: No such file or directory
#15 1309. 34 | #include
#15 1309. …
-
Just wondering if there is way to distribute over CPU (single node, or multi nodes).
It would be very useful features for some sparse models.
-
@HongW2019 could you help to fix third-party-programs-oneCCL.txt according to latest one from oneCCL repo?
xwu99 updated
3 years ago
-
In #2433, CI installs oneCCL from personal GitHub. This should be changed to official repo following oneCCL Gold release.
cc @shirosankaku @mshiryaev
-
From what I understand there are [three ](https://software.intel.com/content/www/us/en/develop/articles/getting-started-with-intel-optimization-of-pytorch.html) separate intel-related improvements to …