[Attention][Spec Decode] Allow independent drafter attention backend selection (#39930)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
Matthew Bonanni
2026-04-27 19:38:09 -07:00
committed by GitHub
parent 146f44b77d
commit fd74c90d9c
6 changed files with 136 additions and 11 deletions
@@ -427,3 +427,90 @@ def test_per_head_quant_scales_backend_selection(
use_per_head_quant_scales=True,
)
assert backend_name in str(exc_info.value)
@pytest.mark.parametrize(
"backend_name,use_non_causal,should_succeed",
[
("FLASH_ATTN", True, True), # FlashAttn supports non-causal
("FLASH_ATTN", False, True), # FlashAttn also works with causal
("FLASHINFER", True, False), # FlashInfer does not support non-causal
("FLASHINFER", False, True), # FlashInfer works with causal
],
)
def test_non_causal_backend_selection(
backend_name: str, use_non_causal: bool, should_succeed: bool
):
"""Test that use_non_causal on AttentionConfig controls backend filtering.
DFlashProposer sets use_non_causal=True on the draft model's
AttentionConfig so only non-causal-capable backends are selected.
The target model keeps use_non_causal=False (default) and can use
any backend.
"""
_cached_get_attn_backend.cache_clear()
attention_config = AttentionConfig(
backend=AttentionBackendEnum[backend_name],
use_non_causal=use_non_causal,
)
cache_config = CacheConfig(block_size=16)
vllm_config = VllmConfig(
attention_config=attention_config, cache_config=cache_config
)
if CudaPlatform is None:
pytest.skip("CudaPlatform not available")
with (
set_current_vllm_config(vllm_config),
patch("vllm.platforms.current_platform", CudaPlatform()),
):
if should_succeed:
backend = get_attn_backend(
head_size=128,
dtype=torch.float16,
kv_cache_dtype=None,
)
assert backend.get_name() == backend_name
else:
with pytest.raises(ValueError) as exc_info:
get_attn_backend(
head_size=128,
dtype=torch.float16,
kv_cache_dtype=None,
)
assert "non-causal" in str(exc_info.value).lower()
def test_non_causal_autoselect_backend():
"""Test that when backend=None with use_non_causal=True, auto-selection
picks a compatible backend.
This simulates the DFlash scenario where the user doesn't specify
--attention-backend or --speculative-config.attention_backend.
The drafter inherits backend=None and auto-selects a backend that
supports non-causal attention.
"""
_cached_get_attn_backend.cache_clear()
attention_config = AttentionConfig(
backend=None,
use_non_causal=True,
)
cache_config = CacheConfig(block_size=16)
vllm_config = VllmConfig(
attention_config=attention_config, cache_config=cache_config
)
if CudaPlatform is None:
pytest.skip("CudaPlatform not available")
with (
set_current_vllm_config(vllm_config),
patch("vllm.platforms.current_platform", CudaPlatform()),
):
backend = get_attn_backend(
head_size=128,
dtype=torch.float16,
kv_cache_dtype=None,
)
assert backend.supports_non_causal()
+3
View File
@@ -54,6 +54,9 @@ class AttentionConfig:
use_fp4_indexer_cache: bool = False
"""If set, use fp4 indexer cache for dsv32 family model (not support yet)"""
use_non_causal: bool = False
"""Whether to use non-causal (bidirectional) attention."""
def compute_hash(self) -> str:
"""
Provide a hash that uniquely identifies all the configs
+15 -1
View File
@@ -5,7 +5,7 @@ import ast
import copy
from typing import TYPE_CHECKING, Any, Literal, get_args
from pydantic import Field, SkipValidation, model_validator
from pydantic import Field, SkipValidation, field_validator, model_validator
from typing_extensions import Self
from vllm.config import LoadConfig
@@ -17,6 +17,7 @@ from vllm.logger import init_logger
from vllm.transformers_utils.config import get_hf_text_config
from vllm.utils.hashing import safe_hash
from vllm.utils.import_utils import LazyLoader, has_arctic_inference
from vllm.v1.attention.backends.registry import AttentionBackendEnum
if TYPE_CHECKING:
from transformers import PretrainedConfig
@@ -106,6 +107,10 @@ class SpeculativeConfig:
inherits the target model's `--moe-backend` setting. Useful when the
drafter and generator require different MoE kernels (e.g. quantized
generator with unquantized drafter)."""
attention_backend: AttentionBackendEnum | None = None
"""Attention backend to use for the draft model. When `None`, the backend is
automatically selected. Useful when the drafter requires a different attention
backend (e.g. DFlash needs a non-causal-capable backend like FLASH_ATTN)."""
max_model_len: int | None = Field(default=None, ge=1)
"""The maximum model length of the draft model. Used when testing the
ability to skip speculation for some sequences."""
@@ -911,6 +916,15 @@ class SpeculativeConfig:
return draft_parallel_config
@field_validator("attention_backend", mode="before")
@classmethod
def _parse_attention_backend(cls, value: Any) -> Any:
if isinstance(value, str):
if value.lower() == "auto":
return None
return AttentionBackendEnum[value.upper()]
return value
@model_validator(mode="after")
def _verify_args(self) -> Self:
if self.tensor_parallel_size is not None:
+1 -6
View File
@@ -81,11 +81,6 @@ def get_attn_backend(
else:
block_size = None
speculative_config = vllm_config.speculative_config
use_non_causal = (
speculative_config is not None and speculative_config.method == "dflash"
)
attn_selector_config = AttentionSelectorConfig(
head_size=head_size,
dtype=dtype,
@@ -97,7 +92,7 @@ def get_attn_backend(
use_mm_prefix=use_mm_prefix,
use_per_head_quant_scales=use_per_head_quant_scales,
attn_type=attn_type or AttentionType.DECODER,
use_non_causal=use_non_causal,
use_non_causal=vllm_config.attention_config.use_non_causal,
use_batch_invariant=envs.VLLM_BATCH_INVARIANT,
)
+12
View File
@@ -1,6 +1,7 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
from dataclasses import replace
from typing import Any
import torch
@@ -67,6 +68,17 @@ class DFlashProposer(SpecDecodeBaseProposer):
# For DFlash we use the input embeddings to embed the mask token
self.parallel_drafting_hidden_state_tensor = None
@override
def _create_draft_vllm_config(self) -> VllmConfig:
base = super()._create_draft_vllm_config()
return replace(
base,
attention_config=replace(
base.attention_config,
use_non_causal=True,
),
)
@override
def _raise_if_multimodal(self):
# Override to allow multimodal inputs since DFlash supports Qwen3.5 models
+18 -4
View File
@@ -1276,15 +1276,29 @@ class SpecDecodeBaseProposer:
Subclasses may override to apply additional config changes.
"""
spec_cfg = self.speculative_config
base = self.vllm_config
if spec_cfg.moe_backend is not None:
return replace(
self.vllm_config,
base = replace(
base,
kernel_config=replace(
self.vllm_config.kernel_config,
base.kernel_config,
moe_backend=spec_cfg.moe_backend,
),
)
return self.vllm_config
# Note (matt): Never inherit the attention backend from base, because there are
# many opportunities for incompatibility, so we always independently autoselect
# unless explicitly specified in the speculative config.
base = replace(
base,
attention_config=replace(
base.attention_config,
backend=spec_cfg.attention_backend,
),
)
return base
def _get_model(self) -> nn.Module:
"""