Adds support for allgather_into_tensor_coalesced
and `reduce_scatte…
#348
Job | Run time |
---|---|
9m 32s | |
6m 51s | |
16m 23s |
allgather_into_tensor_coalesced
and `reduce_scatte…
#348
Job | Run time |
---|---|
9m 32s | |
6m 51s | |
16m 23s |