transformers
d9183573 - Merge branch 'main' into fix-empty-fsdp-shards

Commit

5 days ago

Merge branch 'main' into fix-empty-fsdp-shards

References

#48204 - [Umbrella] Enable FSDP2 + expert parallelism via a 2-D (fsdp, tp) device mesh

#48237 - Handle empty FSDP shards in sharded loading: crash and hang fix

Author

qgallouedec

qgallouedec

Parents

Loading