[Perf][DSV4.1] Remove MegaMoE padding and shared padding workaround (#57204)

Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
This commit is contained in:
Woosuk Kwon
2026-09-16 19:09:30 +00:00
committed by GitHub
parent bc0f47cd03
commit 30b847c1fa
2 changed files with 42 additions and 157 deletions
+37 -92
View File
@@ -300,8 +300,10 @@ def test_deepseek_v4_mega_moe_weight_loader_uses_ep_expert_ownership():
@pytest.mark.parametrize("intermediate_size", [512, 2304])
def test_deepseek_v4_mega_moe_padding_preserves_weights(monkeypatch, intermediate_size):
"""Pad gate/up separately and zero every added down-projection column."""
def test_deepseek_v4_mega_moe_preserves_checkpoint_dimensions(
monkeypatch, intermediate_size
):
"""Keep native widths so V4.1 routed and shared experts can fuse."""
experts = DeepseekV4MegaMoEExperts(
SimpleNamespace(
scheduler_config=SimpleNamespace(max_num_batched_tokens=4),
@@ -333,29 +335,15 @@ def test_deepseek_v4_mega_moe_padding_preserves_weights(monkeypatch, intermediat
)
experts.finalize_weights()
padded_size = experts.intermediate_size
assert padded_size == (intermediate_size + 511) // 512 * 512
for actual, original in zip(experts._transformed_l1_weights, originals[:2]):
assert experts.intermediate_size == intermediate_size
transformed = (*experts._transformed_l1_weights, *experts._transformed_l2_weights)
for actual, original in zip(transformed, originals):
expected = (
original
if actual.dtype == torch.int8
else (experts._ue8m0_uint8_to_float(original))
else experts._ue8m0_uint8_to_float(original)
)
actual = actual.view(expected.dtype).unflatten(1, (2, padded_size))
assert torch.equal(
actual[:, :, :intermediate_size],
expected.unflatten(1, (2, intermediate_size)),
)
assert torch.count_nonzero(actual[:, :, intermediate_size:]) == 0
for actual, original in zip(experts._transformed_l2_weights, originals[2:]):
expected = (
original
if actual.dtype == torch.int8
else (experts._ue8m0_uint8_to_float(original))
)
actual = actual.view(expected.dtype)
assert torch.equal(actual[..., : original.shape[-1]], expected)
assert torch.count_nonzero(actual[..., original.shape[-1] :]) == 0
assert torch.equal(actual.view(expected.dtype), expected)
transformed_l1 = experts._transformed_l1_weights
experts.finalize_weights()
@@ -363,20 +351,19 @@ def test_deepseek_v4_mega_moe_padding_preserves_weights(monkeypatch, intermediat
@pytest.mark.parametrize(
"hidden_size,intermediate_size,block_size,fused,mxfp8",
"hidden_size,intermediate_size,block_size,mxfp8",
[
pytest.param(128, 512, 128, True, False, id="aligned-block128"),
pytest.param(128, 512, 128, True, True, id="aligned-block128-mxfp8"),
pytest.param(5120, 2304, 32, True, False, id="deepseek-v41-flash"),
pytest.param(128, 2304, 128, True, False, id="padded-block128"),
pytest.param(5120, 2304, 32, True, True, id="deepseek-v41-flash-mxfp8"),
pytest.param(128, 2304, 128, False, False, id="padded-packed-scale-fallback"),
pytest.param(128, 512, 128, False, id="aligned-block128"),
pytest.param(128, 512, 128, True, id="aligned-block128-mxfp8"),
pytest.param(5120, 2304, 32, False, id="deepseek-v41-flash"),
pytest.param(128, 2304, 128, False, id="native-block128"),
pytest.param(5120, 2304, 32, True, id="deepseek-v41-flash-mxfp8"),
],
)
def test_deepseek_v4_mega_moe_finalizes_native_shared_expert_weights(
monkeypatch, hidden_size, intermediate_size, block_size, fused, mxfp8
monkeypatch, hidden_size, intermediate_size, block_size, mxfp8
):
"""Shared fusion preserves checkpoint channels and zeros padded channels."""
"""Shared fusion preserves checkpoint weights and scales at native widths."""
class FakeDeepGemm:
transformed_dims: list[tuple[int, int]] = []
@@ -476,12 +463,6 @@ def test_deepseek_v4_mega_moe_finalizes_native_shared_expert_weights(
originals = []
for linear in (shared_experts.gate_up_proj, shared_experts.down_proj):
scale = linear.weight_scale if mxfp8 else linear.weight_scale_inv
if not fused:
# Already packed scales cannot be padded as checkpoint UE8M0 bytes.
linear.weight_scale_inv = scale_parameter(
linear.weight.shape[0], linear.weight.shape[1] // 128
)
continue
scale.data.view(torch.uint8).random_(124, 130)
block_m, block_k = linear.weight_block_size
scale_1x32 = (
@@ -493,70 +474,34 @@ def test_deepseek_v4_mega_moe_finalizes_native_shared_expert_weights(
monkeypatch.setattr("vllm.utils.deep_gemm._import_deep_gemm", lambda: FakeDeepGemm)
original_gate_up_ptr = shared_experts.gate_up_proj.weight.data_ptr()
original_down_ptr = shared_experts.down_proj.weight.data_ptr()
experts.finalize_weights(shared_experts)
assert experts.has_fused_shared_experts is fused
if not fused:
assert experts.intermediate_size == 2560
assert FakeDeepGemm.transformed_dims == [(3, 3)]
assert experts.num_shared_experts == 0
assert shared_experts.gate_up_proj.weight.data_ptr() == original_gate_up_ptr
assert shared_experts.down_proj.weight.data_ptr() == original_down_ptr
assert shared_experts.gate_up_proj.weight.shape == (
2 * intermediate_size,
hidden_size,
)
assert shared_experts.down_proj.weight.shape == (hidden_size, intermediate_size)
return
assert experts.has_fused_shared_experts
assert experts.intermediate_size == intermediate_size
assert FakeDeepGemm.transformed_dims == [(3, 3), (2, 2)]
padded_size = experts.intermediate_size
assert FakeDeepGemm.scale_inputs[-2:] == [
(1, 2 * padded_size, hidden_size // 32),
(1, hidden_size, padded_size // 32),
(1, 2 * intermediate_size, hidden_size // 32),
(1, hidden_size, intermediate_size // 32),
]
for actual, original, fill in (
for transformed, actual_scale, (weight, scale) in zip(
(
experts._transformed_shared_l1_weights[0].view(torch.uint8),
originals[0][0],
0,
experts._transformed_shared_l1_weights,
experts._transformed_shared_l2_weights,
),
(FakeDeepGemm.scale_values[-2][0], originals[0][1], 1),
FakeDeepGemm.scale_values[-2:],
originals,
):
actual = actual.unflatten(0, (2, padded_size))
assert torch.equal(
actual[:, :intermediate_size], original.unflatten(0, (2, intermediate_size))
)
assert torch.all(actual[:, intermediate_size:] == fill)
for actual, original, fill in (
(
experts._transformed_shared_l2_weights[0].view(torch.uint8),
originals[1][0],
0,
),
(FakeDeepGemm.scale_values[-1][0], originals[1][1], 1),
):
assert torch.equal(actual[:, : original.shape[1]], original)
assert torch.all(actual[:, original.shape[1] :] == fill)
if padded_size != intermediate_size:
# Generic linear post-load hooks still receive checkpoint-shaped weights.
for linear, (weight, _) in zip(
(shared_experts.gate_up_proj, shared_experts.down_proj), originals
):
assert torch.equal(linear.weight.view(torch.uint8), weight)
assert shared_experts.gate_up_proj.weight.data_ptr() == original_gate_up_ptr
assert shared_experts.down_proj.weight.data_ptr() == original_down_ptr
else:
assert shared_experts.gate_up_proj.weight.data_ptr() != original_gate_up_ptr
assert (
experts._transformed_shared_l1_weights[0].data_ptr()
== shared_experts.gate_up_proj.weight.data_ptr()
)
assert (
experts._transformed_shared_l2_weights[0].data_ptr()
== shared_experts.down_proj.weight.data_ptr()
)
assert torch.equal(transformed[0].view(torch.uint8), weight)
assert torch.equal(actual_scale[0], scale)
assert shared_experts.gate_up_proj.weight.data_ptr() != original_gate_up_ptr
assert (
experts._transformed_shared_l1_weights[0].data_ptr()
== shared_experts.gate_up_proj.weight.data_ptr()
)
assert (
experts._transformed_shared_l2_weights[0].data_ptr()
== shared_experts.down_proj.weight.data_ptr()
)
transformed_l1 = experts._transformed_shared_l1_weights
experts.finalize_weights(shared_experts)