Skip to content

ZeRO1: Add bucketting logic to control the size of tensors for all-gather/reduce-scatter #6540

ZeRO1: Add bucketting logic to control the size of tensors for all-gather/reduce-scatter

ZeRO1: Add bucketting logic to control the size of tensors for all-gather/reduce-scatter #6540

Annotations

1 warning

The logs for this run have expired and are no longer available.