[EPLB] Nixl communicator optimization. Zero-copy transfers (#41633)

Signed-off-by: ilmarkov <markovilya197@gmail.com>
Signed-off-by: Markov Ilya <markovilya19@gmail.com>
Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: Markov Ilya <markovilya19@gmail.com>
Co-authored-by: Nicolò Lucchesi <nlucches@redhat.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
This commit is contained in:
Ilya Markov
2026-06-04 03:40:34 +00:00
committed by GitHub
co-authored by Markov Ilya Nicolò Lucchesi Tyler Michael Smith Tyler Michael Smith
parent f0cd590d62
commit 4f423bd5bc
12 changed files with 287 additions and 219 deletions
+4 -1
View File
@@ -1287,10 +1287,12 @@ def _test_body_eplb(
expert_weights = [list(eplb_moe_layer.get_expert_weights())]
expert_buffer = [torch.empty_like(w) for w in expert_weights[0]]
communicator = create_eplb_communicator(
group_coordinator=get_eplb_group(),
backend=vllm_config.parallel_config.eplb_config.communicator,
expert_weights=expert_weights[0],
expert_weights=expert_weights,
expert_buffer=expert_buffer,
)
# Rearrange expert weights across EP ranks
@@ -1298,6 +1300,7 @@ def _test_body_eplb(
old_global_expert_indices=initial_indices.unsqueeze(0),
new_global_expert_indices=shuffled_indices.unsqueeze(0),
expert_weights=expert_weights,
expert_buffer=expert_buffer,
ep_group=cpu_group,
communicator=communicator,
)