shard

★ 0 updated 17d ago ⑂ fork

Pipeline-parallel LLM inference across GPUs on separate machines.

No plain-English explanation yet — one is being written right now. Check back in a minute.