[Bugfix][Spec Decode] DSpark: inherit the target's attention backend when the speculative config names none (#52288)
This commit is contained in:
@@ -21,12 +21,18 @@ def load_dspark_model(target_model: nn.Module, vllm_config: VllmConfig) -> nn.Mo
|
|||||||
from vllm.model_executor.models.qwen3_dflash import dflash_has_any_non_causal
|
from vllm.model_executor.models.qwen3_dflash import dflash_has_any_non_causal
|
||||||
from vllm.model_executor.models.utils import get_draft_quant_config
|
from vllm.model_executor.models.utils import get_draft_quant_config
|
||||||
|
|
||||||
|
# None re-runs backend auto-selection for the draft, which can pick a
|
||||||
|
# different attention class than the target; fall back to the target's.
|
||||||
|
draft_attention_backend = (
|
||||||
|
speculative_config.attention_backend or vllm_config.attention_config.backend
|
||||||
|
)
|
||||||
|
|
||||||
draft_vllm_config = replace(
|
draft_vllm_config = replace(
|
||||||
vllm_config,
|
vllm_config,
|
||||||
attention_config=replace(
|
attention_config=replace(
|
||||||
vllm_config.attention_config,
|
vllm_config.attention_config,
|
||||||
use_non_causal=dflash_has_any_non_causal(draft_model_config.hf_config),
|
use_non_causal=dflash_has_any_non_causal(draft_model_config.hf_config),
|
||||||
backend=speculative_config.attention_backend,
|
backend=draft_attention_backend,
|
||||||
),
|
),
|
||||||
cache_config=(
|
cache_config=(
|
||||||
replace(
|
replace(
|
||||||
|
|||||||
Reference in New Issue
Block a user