Fixed logging level in distributed mode (#4284) 771ac6c unverified imyhxy fkwong commited on Aug 3, 2021
Update `profile()` for CUDA Memory allocation (#4239) d8f1883 unverified glenn-jocher commited on Jul 30, 2021
Update DDP for `torch.distributed.run` with `gloo` backend (#3680) fad27c0 unverified glenn-jocher commited on Jun 19, 2021
Merge `develop` branch into `master` (#3518) f3c3d2c unverified glenn-jocher Piotr Skalski SkalskiP Peretz Cohen tudoulei chocosaj BuildTools developer0hye Sam_S Samridha Shrestha edificewang commited on Jun 8, 2021
check `batch_size % utilized_device_count` (#3276) f340235 unverified glenn-jocher commited on May 21, 2021
Update git_describe() for remote dir usage (#2606) 77415a4 unverified glenn-jocher commited on Mar 26, 2021
PyTorch Hub models default to CUDA:0 if available (#2472) 9b11f0c unverified glenn-jocher commited on Mar 15, 2021
Fix torch multi-GPU --device error (#1701) 035ac82 unverified Nanobit glenn-jocher commited on Dec 16, 2020
FLOPS computation device bug fix (#1447) 05a955a unverified yujun glenn-jocher commited on Nov 19, 2020
PyTorch version to screen and cleanup (#1325) 19e2482 unverified glenn-jocher commited on Nov 9, 2020
Fix redundant outputs via Logging in DDP training (#500) 4949401 unverified Nanobit glenn-jocher commited on Aug 11, 2020